The Technology Expo at CASCON 2017 will provide an excellent opportunity to highlight your research results and generate interest in upcoming product and product areas. Exhibits on emerging technologies are encouraged. The format of an exhibit may be a demonstration, poster presentation, or a short talk. A booth is provided at no cost to the exhibitors. Each demonstration booth will be set up with power and wireless internet access. Exhibitors are responsible for bringing the necessary equipment such as machines, laptops, monitors, power cables, and security locks; and any printed materials including posters.

Click here to download Call for Exhibits

Accepted exhibits

Congratulations to the exhibit chairs and thank you for your submissions. See you all at CASCON 2017.

See details below:

Exhibit Sessions - 11:45 to 12:45 and 17:00 to 19:30

  • P1: A Machine Learning Based Application Autonomic Management System on Software Defined Networks
    Booth Number: P1
    Title: A Machine Learning Based Application Autonomic Management System on Software Defined Networks
    Chair(s): Nasim Beigi , MarK Shtern , Marin Litoiu , Joe Wigglesworth, and Radu Mateescu
    Format: Poster
    Description: In this work, we propose and implement a machine learning based autonomic manager that controls the bandwidth rates allocated to each scenario of a web application to postpone scaling out for as long as possible despite change in the workload and application configuration. Through experiments, we demonstrate that the autonomic manager is able to quickly meet Service level Agreement (SLA) and reduce the SLA violations by 56% compared to a previous heuristic-based approach.
    Industry Company: IBM
    Project Number 1004
    Product Name: IBM PureApp
    Research Team: York University: Nasim Beigi, Mark Shtern, Marin Litoiu 
    IBM CAS: Joe Wigglesworth, and Radu Mateescu 

  • D1: Analytics for Endpoints System Data
    Booth Number: D1
    Title: Analytics for Endpoints System Data
    Chair(s): arash Habibi Lashkari
    Format: Demo
    Description: There is a hypothesis that by collecting large amounts of data from endpoints and doing analysis, comparisons and data mining a system could detect systems infected with malware that is currently undetectable by state of the art security systems. This same data could be used for finding eavesdropping software used by a rogue employee and event data theft. At this point we present the first part of the User's Universal Security Profile namely network profile out of all five profiles.
    Industry Company: IBM
    Project Number 1042
    Product Name: UUP
    Web Site: http://unb.ca/cic/research/applications.html
    Research Team: Canadian Institute for Cybersecurity (CIC): Arash Habibi Lashkari 

  • P2: Approximation Algorithms for Instruction Scheduling and Rapid Prototyping in Coconut
    Booth Number: P2
    Title: Approximation Algorithms for Instruction Scheduling and Rapid Prototyping in Coconut
    Chair(s): Curtis D'Alves
    Format: Poster
    Description: We explore approximation algorithm based solutions for instruction scheduling on the IBM Z architecture and MASS API capable of scheduling near-optimal code. This includes a relaxation of the problem to a continuous model and a stochastic non-linear programming solution. In collaboration with this project, we have extended our development environment Coconut for modelling code dependencies and generation of scheduled code.
    Industry Company: IBM
    Project Number 1006
    Product Name: Approximation Algorithms for Instruction Scheduling
    Research Team: McMaster University: Curtis D'Alves  Dr Christopher Kumar Anand  Wolfram Kahl 
    IBM: Robert Enenkel  Bill O'Farrell 

  • P3: Bridging the Gap between Deep Learning and Sparse Matrix Format Selection
    Booth Number: P3
    Title: Bridging the Gap between Deep Learning and Sparse Matrix Format Selection
    Chair(s): Yue Zhao
    Format: Poster
    Description: Sparse matrix vector multiplication (SpMV) is an important kernel in many applications. In this work, we conduct a systematic exploration on the promise and challenges of deep learning for the sparse matrix format selection. We propose a set of novel techniques to solve special challenges to deep learning, including input matrix representations, a late-merging deep neural network structure design, and the use of transfer learning to alleviate cross-architecture portability issues.
    Industry Company: IBM
    Project Number 1009
    Product Name: Cognitive Computing-Based Compilation
    Research Team: NCSU: Yue Zhao, Xipeng Shen 
    IBM: Graham Yiu 

  • P4: Center for Health Informatics and Analytics
    Booth Number: P4
    Title: Center for Health Informatics and Analytics
    Chair(s): Randy Giffen
    Format: Poster
    Description: The Center for Health Informatics and Analytics is a collaboration between Memorial University and IBM Canada. The goal of the center is to jointly develop and validate solutions that help translate knowledge and data into action. The solutions will support both clinical and administrative scenarios. The center is based on a vision for healthcare transformation. Lower healthcare costs can be achieved by improving efficiency and effectiveness and ultimately by improving health.
    Industry Company: IBM
    Product Name: Center for Health Informatics and Analytics
    Web Site: http://www.med.mun.ca/CHIA/Home.aspx
    Research Team: IBM Canada: Randy Giffen 

  • P5: Cold Object Segregation
    Booth Number: P5
    Title: Cold Object Segregation
    Chair(s): Scott Young , Kenneth Kent , Gerhard Dueck
    Format: Poster
    Description: Cold (infrequently referenced) objects can take up extra space in the areas of memory reserved for frequently accessed, tenured objects. Identifying all cold objects in tenured space during runtime is expensive. This project seeks to find an algorithm for selecting candidate cold objects to be moved to secondary storage so that the difference between time saved during garbage collection and time spent performing the selection is maximized.
    Industry Company: IBM
    Project Number 1001
    Product Name: OMR and J9
    Web Site: http://www.unb.ca/research/casatlantic/
    Research Team: UNB: Scott Young  Kenneth Kent  Gerhard Dueck 
    IBM Canadda: Charlie Gracie 

  • D2: Eliminating the societal impact of learning disabilities with cognitive computing
    Booth Number: D2
    Title: Eliminating the societal impact of learning disabilities with cognitive computing
    Chair(s): Devan Shah , Will Scott , Larry Lindsay
    Format: Demo
    Description: Nearly one billion individuals live with some form of learning disability, or benefit from learning in different ways, such as those living with autism spectrum disorder. What if machine learning and cognitive computing-based assistive technologies could eliminate the societal and socio-economic challenges faced by these individuals in the future? In this hands-on demo, the audience will learn about IBMs cutting-edge research in this space, the IBM AbilityLab Content Clarifier.
    Web Site: http://contentclarifier.mybluemix.net/
    Research Team: IBM: Devan Shah  Will Scott  Larry Lindsay  Thomas Brunet 

  • D3: Foodie: A Conversational Agent for the Smart Kitchen
    Booth Number: D3
    Title: Foodie: A Conversational Agent for the Smart Kitchen
    Chair(s): Prashanti Angara , Miguel Jimenez , Hausi Muller , Ulrike Stege
    Format: Demo
    Description: Conversational agents aim to offer an alternative to traditional methods for humans to engage with technology. Foodie is a conversational kitchen assistant that uses IBM Watsons conversational services to recognize users intents and understand events related to the users and their context. Foodie uses services from CAPRecipesour context-aware personalized recipe recommender system, SmarterContextour personal context management system, and selected publicly available nutrition databases.
    Industry Company: University of Victoria
    Product Name: Foodie Fooderson
    Research Team: University Of Victoria: Prashanti Priya Angara  Miguel Jimenez  Hausi Muller  Ulrike Stege 
    IBM Canada Ltd.: Joanna W. Ng 

  • D4: IBM Data Science Experience Local
    Booth Number: D4
    Title: IBM Data Science Experience Local
    Chair(s): Mihai Iacob , Mihai Nicolae
    Format: Demo
    Description: Bring Data Science Experience (DSX) to your private cloud and use the tools you know and love: Jupyter notebooks, Zeppelin notebooks, RStudio, all powered by Apache Spark. The admin dashboard lets you control your cluster from managing users to CPU, memory, network, and storage usage. DSX provides a collaborative environment where a team can collaborate on a project, and share notebooks, data sources, etc.
    Industry Company: IBM
    Business unit IBM Analytics
    Product Name: IBM Data Science Experience Local
    Web Site: https://datascience.ibm.com/local
    Research Team: IBM Canada: Mihai Iacob  Mihai Nicolae 

  • P8: IBM High-Performance Mathematical Function Libraries
    Booth Number: P8
    Title: IBM High-Performance Mathematical Function Libraries
    Chair(s): Robert Enenkel
    Format: Poster
    Description: We show how to accelerate numerically intensive applications with the IBM MASS and ATLAS high-performance math libraries, which are tuned for optimal performance on IBM POWER and zSeries processors. MASS provides elementary functions (e.g. exp, log, sin, cos, etc.) while ATLAS provides linear algebra functions (e.g. DGEMM). We also show how to use MASS automatically through the IBM XL C/C++ and Fortran compilers.
    Business unit IBM Systems
    Product Name: IBM XL compilers, MASS, ATLAS
    Web Site: http://www-03.ibm.com/software/products/en/mathacc
    Research Team: IBM: Robert Enenkel,  Research Commercialization Leader 
    McMaster University: Prof. Christopher Anand (CAS Fellow)  Curtis d'Alves (CAS Student) 

  • P9: MASS library scalar math functions for the Go Language
    Booth Number: P9
    Title: MASS library scalar math functions for the Go Language
    Chair(s): Bill O'Farrell
    Format: Poster
    Description: Improving scalar math performance in the Go programming language on IBM Z. We took the scalar MASS libraries and translated them from Z assembly to Go assembly, had them approved as open source, and contributed them to Go. The translation was done automatically (mostly) via a python script. The project is complete and was included in go 1.9.
    Industry Company: IBM
    Business unit Systems
    Product Name: Go Language (Open-Source) on IBM Z
    Research Team: IBM: Bill O'Farrell  Kang Zhang  Annita Zhang  Robert Enenkel 

  • P6: NUMA GC
    Booth Number: P6
    Title: NUMA GC
    Chair(s): Maria Patrou
    Format: Poster
    Description: Non-uniform Memory Access (NUMA) systems use different types of memory accesses. IBM's Java Virtual Machine identifies a NUMA architecture, while using memory and threads from the available nodes in a distributed way. A design for a node-isolated memory achieved by a node-heap resize functionality and a thread policy is proposed. Finally, different modes regarding hardware and thread characteristics are investigated to identify the application attributes that can benefit from specific modes.
    Industry Company: IBM
    Project Number 1020
    Product Name: IBM SDK
    Web Site: http://www.unb.ca/research/casatlantic/
    Research Team: University of New Brunswick: Maria Patrou  Kenneth B. Kent  Gerhard W. Dueck 
    IBM Canada: Charlie Gracie  Aleksandar Micic 

  • D5: Optimization of Parametric GPU kernel
    Booth Number: D5
    Title: Optimization of Parametric GPU kernel
    Chair(s): Marc Moreno Maza , Haoze Yuan , Masoud Ataei Jaliseh
    Format: Demo
    Description: This work deals with the optimization of computer programs targeting Graphics Processing Units (GPUs). The goal is to lift, from programmers to optimizing compilers, the heavy burden of determining program details that are dependent on the hardware characteristics. The expected benefit is to improve robustness, portability and efficiency of the generated computer programs.
    Industry Company: IBM
    Project Number 880
    Product Name: IBM XL compiler
    Web Site: www.metafork.org
    Research Team: University of Western Ontarioo: Marc Moreno-Maza  Haoze Yuan  Masoud Ataei 
    IBM Canada: Jeeva Paudel 

  • D6: Optimizing your Workforce by Leveraging IBM Predictive Analytics
    Booth Number: D6
    Title: Optimizing your Workforce by Leveraging IBM Predictive Analytics
    Chair(s): Rebecca Young
    Format: Demo
    Description: IBM Predictive Analytics is used to identify how to address unwanted employee attrition, a problem that can affect any organization within any industry. This solution can be deployed at any stage of the data analytics maturity curve and users from both business and technical groups can leverage IBM Predictive Analytics to derive value and address their organizational challenges.
    Industry Company: IBM
    Business unit Analytics
    Product Name: IBM Predictive Analytics
    Research Team: IBM: Rebecca Young 

  • D7: PANORAMA: Deployment Specification in the Context of Large Scale Systems
    Booth Number: D7
    Title: PANORAMA: Deployment Specification in the Context of Large Scale Systems
    Chair(s): Miguel Jimenez , Hausi Muller , Gabriel Tamura , Norha M. Villegas
    Format: Demo
    Description: DevOps has dramatically influenced how/when deployment is realised. However, little effort has been devoted to the need for deployment and configuration specifications to support the various levels of detail and abstraction present in large-scale systems. In this exhibit, we present PANORAMA, our envisioned environment to design and visualise large-scale continuous deployment based on the OASIS TOSCA specification.
    Research Team: University of Victoria: Miguel Jimnez, Hausi Mller 
    Universidad Icesi: Gabriel Tamura, Norha Villegas 

  • P7: Phishing Numbers
    Booth Number: P7
    Title: Phishing Numbers
    Chair(s): Guy Jourdan , Qian Cui
    Format: Poster
    Description: Are popular websites hacked to host phishing attacks? Why do phishing sites sometimes redirect their victims to other sites? How often are phishing attacks hosted on compromised servers? What are the most common paths used in phishing attacks?... We have been gathering sample phishing attacks for the past 18 months, and have conducted a number of analyses to answer some of these questions. This exhibit will provide an overview of our findings to date.
    Industry Company: IBM
    Project Number 921
    Product Name: Security Forensics
    Research Team: University of Ottawa: Guy-Vincent Jourdan  Gregor Bochmann  Qian Cui  Sophie LePage 
    IBM: Vio Onut  Russ Couturier 

  • P10: Poly1305 acceleration on the IBM z14
    Booth Number: P10
    Title: Poly1305 acceleration on the IBM z14
    Chair(s): Bill O'Farrell , Umme Salma Gadriwala
    Format: Poster
    Description: IBM z14 has a new integer multiply instruction which is very useful for use in cryptography. Implemented with the floating point pipeline, the instruction -- Vector Multiply Sum Logical -- can be used to accelerate "limbified" multiplication operations that are common in cryptography. We explored this instruction by applying it to a a high-performance version of the Poly1305 signing algorithm.
    Industry Company: IBM
    Business unit Systems
    Product Name: Go Language (Open-Source) on IBM Z
    Research Team: McMaster University: Umme Salma Gadriwala  Christopher Kumar Anand  Curtis DAlves 
    IBM: Bill O'Farrell  Jonathan Bradbury  Michael Munday 

  • P11: Quantifying Duplication to Improve Data Quality
    Booth Number: P11
    Title: Quantifying Duplication to Improve Data Quality
    Chair(s): Yu Huang
    Format: Poster
    Description: We present a record deduplication framework that differentiates terms during the matching process to improve overall accuracy. We also define a duplication metric that quantifies the level of duplication for an attribute value, and within an attribute. This metric can be used by analysts to understand the distribution and similarity of values during the data cleaning process.
    Industry Company: IBM
    Project Number 923
    Product Name: IBM Watson Analytics
    Research Team: McMaster University: Yu Huang  Fei Chiang 
    IBM Ottawa: Martin Petitclerc 
    IBM Germany: Albert Maier  Yannick Saillet 
    IBM Chicago: Damir Spisic 
    IBM Toronto: Calisto Zuzarte 

  • P12: Recommender System for IBM Watson Analytics-Algorithm Development
    Booth Number: P12
    Title: Recommender System for IBM Watson Analytics-Algorithm Development
    Chair(s): Parisa Lak
    Format: Poster
    Description: IBM Watson Analytics (WA) provides users with visualization recommendations for their provided dataset using a rule based system. Recommender systems (RS) are a set of learning algorithms that provide personalized recommendations to the users based on their historical preferences. In this work, we design a RS that receives users historical interaction with WA as input, and provides a list of recommendations that are uniquely generated based on his/her prior interaction with the system.
    Industry Company: IBM
    Project Number 919
    Product Name: Watson Analytics
    Research Team: Ryerson University: Parisa Lak  Can Kavaklioglu  Ayse Bener  Andriy Miranskyy 
    IBM: Martin Petitclerc  Graham Wills 

  • P13: Runtime Automatic Parallelization of JVM Applications
    Booth Number: P13
    Title: Runtime Automatic Parallelization of JVM Applications
    Chair(s): Akihiro Hayashi , Gita Koblents
    Format: Poster
    Description: In this IBM-Rice University collaboration, we push the bounds on integrating JVM applications with GPU accelerators. This work uses advanced code generation and performance prediction techniques to effectively and automatically offload JVM applications including Apache Spark to hardware accelerators.
    Industry Company: IBM
    Project Number 900
    Product Name: IBM Testarossa Compiler, IBM J9 VM, IBM Spark-DK
    Research Team: Rice University: Akihiro Hayashi  Max Grossman  Vivek Sarkar 
    IBM Canada: Gita Koblents  Jimmy Kwa  Alon Shalev Housfater 
    IBM Research - Tokyo: Kazuaki Ishizaki 

  • P14: Supporting Microservice Evolution
    Booth Number: P14
    Title: Supporting Microservice Evolution
    Chair(s): Julia Rubin
    Format: Poster
    Description: Microservices have become a popular pattern for deploying scale-out application logic. An advantage of using microservices is their loose coupling, which leads to agile and rapid evolution, and continuous re-deployment. However, developers are tasked with managing this evolution and largely do so manually by continuously evaluating low-level service behaviors. This is tedious, error-prone, and slow. In this exhibit, we discuss how our approach can help engineers with these tasks.
    Industry Company: IBM
    Project Number 1040
    Product Name: IBM Bluemix
    Research Team: University of British Columbia: Harshavardhan Kadiyala,   Bo Hu,  Ivan Beschastnikh,   Julia Rubin 
    IBM: John Steinbacher,   Tony Erwin  
    Federal University of Pernambuco: Adalberto R. Sampaio Jr.,  Nelson Rosa 

  • P15: ThreadedPaws: A Serious Game for Learning the Pitfalls of Concurrent Programming
    Booth Number: P15
    Title: ThreadedPaws: A Serious Game for Learning the Pitfalls of Concurrent Programming
    Chair(s): Luisa Rojas , Jeremy Bradbury , Michael Miljanovic
    Format: Poster
    Description: Advances in multi-core processors continue to increase the need for concurrent programming. Unfortunately, writing concurrent programs remains difficult due to the many, possibly unexpected program executions. Furthermore, students learning concurrent programming need to comprehend and avoid common pitfalls such as data races and deadlocks. To address this need, we have developed Threaded Paws, a game-based learning tool that teaches students to identify and fix concurrency pitfalls and bugs.
    Research Team: University of Ontario IT: Luisa Rojas  Jeremy S. Bradbury  Michael Miljanovic 

  • P16: Variability-aware analysis of C++ Code
    Booth Number: P16
    Title: Variability-aware analysis of C++ Code
    Chair(s): Nazim Uddin Bhuiyan , Samer AL Masri
    Format: Poster
    Description: IBMs Eclipse OMR project is a C++ language-agnostic library of run-time components. It implements software variability through static polymorphism (specifically, extensible classes) and #ifdef directives. OMR developers need to reason about multiple architectures and language extensions in an efficient manner, but to the best of our knowledge, there are no existing variability-aware C++ analysis tools that support OMRs variability implementation. This project aims to provide such support.
    Industry Company: IBM
    Project Number 1041
    Product Name: Eclipse OMR
    Research Team: University of Alberta: Samer AL Masri  Nazim Uddin Bhuiyan  Sarah Nadi 
    IBM Canada: Matthew Gaudet 


Expo Reception / Short Talks / PechaKucha Expo Presentations

Welcome - 17:15 to 17:45

Short Talk Sessions - 17:45 to 18:15 (2 talks).   Talks order as follows (15 mins each)

  • S1: Introduction to Digital Process Automation
    Booth Number: S1
    Title: Introduction to Digital Process Automation
    Chair(s): Sebastian Carbajales , Allen Chan
    Format: Short Talk
    Description: Mundane, repetitive tasks and inflexible processes can hinder the speed where business can react to new challenges or new opportunities. Come here to learn how IBM Digital Process Automation can increase your business success by using the latest automation and cognitive technology.
    Industry Company: IBM
    Business unit BPM
    Product Name: IBM Digital Process Automation
    Web Site: https://www.ibm.com/cloud-computing/products/digit
    Research Team: IBM: Sebastian Carbajales, IBM BPM Development 
    IBM: Allen Chan, IBM Distinguished Engineer  

  • S2: Safety Assurance in the Automotive Domain and Model Based Techniques to Support it
    Booth Number: S2
    Title: Safety Assurance in the Automotive Domain and Model Based Techniques to Support it
    Chair(s): Sahar Kokaly
    Format: Short Talk
    Description: Software has become a large part of many systems including autonomous vehicles. To address issues such as safety and security, compliance of software has emerged as a key issue. Software compliance is costly and is often accomplished by producing "assurance cases". As systems evolve, maintaining assurance cases multiplies the effort. We propose model-based techniques to address issues such as assurance case evolution and reuse. We focus on the automotive domain and the ISO 26262 safety standard.
    Research Team: McMaster University/University of Toronto: Sahar Kokaly 
    University of Toronto: Rick Salay 
    University of Toronto: Marsha Chechik 
    McMaster University: Mark Lawford 
    McMaster University: Tom Maibaum 

Break - 18:15 to 18:25

PechaKucha Sessions - 18:25 to 19:30. (13 talks).   Presentations order as follows (5 mins each)

  • A Hierarchical and Dynamic Security Control System for Large Scale Systems
    Exhibit Session: Tuesday D1
    Title: A Hierarchical and Dynamic Security Control System for Large Scale Systems
    Chair(s): Yar Rouf , Marin Litoiu
    Format: Demo
    Description: We propose a hierarchical distributed architecture for security control in order to partition responsibility and workload among many security controllers. Our architecture proposes a more simplified way of defining security rules to allow security to be enforced on an operational level, rather than a development level. We demonstrate the ability of our system to expand with additional sensors, traffic data and detection rules at run time in a systematic manner.
    Industry Company: IBM
    Project Number 1004
    Product Name: IBM PureApp
    Research Team: York University: Yar Rouf,   Mark Shtern,   Marios Fokaefs,   Marin Litoiu 
    IBM: Vio Onut 

  • A Scalability-oriented Benchmark Suite for Node.js on the Cloud
    Exhibit Session: Tuesday P1
    Title: A Scalability-oriented Benchmark Suite for Node.js on the Cloud
    Chair(s): Jiapeng Zhu , Panagiotis (Panos) Patros
    Format: Poster
    Description: Clouds provide on-demand and pay-as-you-go computing resources to their end users. Node.js is a popular choice for cloud applications as it is event-driven, asynchronous and non-blocking. However, Node.js underutilizes multi-core CPUs due to its single-threaded nature, which limits its scalability on the cloud. Because no scalability-oriented benchmark suite exists for Node.js, we will focus on developing one. Our benchmarking suite will facilitate and quantify Node.js scalability improvements.
    Industry Company: IBM
    Project Number 1002
    Product Name: IBM SDK for Node.js
    Web Site: http://www.unb.ca/research/casatlantic/
    Research Team: University of New Brunswick: Jiapeng Zhu, Patros Panagiotis, Kenneth Kent 
    IBM Canada: Michael Dawson 

  • A Scalable Architecture for the Internet of Things
    Exhibit Session: Tuesday P2
    Title: A Scalable Architecture for the Internet of Things
    Chair(s): Brian Ramprasad , Marios Fokaefs , Marin Litoiu
    Format: Poster
    Description: As the volume and variety of sensors continues to grow, a solution is needed to process large volumes of streaming data produced by IoT devices. We propose an architecture that is flexible and scalable and an emulation framework to test the scalability of large scale sensor networks. Our design employs IoT gateways using Node-RED, a messaging broker as an aggregation point, and a streaming engine to pre-process and persist the IoT data.
    Industry Company: IBM
    Project Number 1004
    Product Name: IBM Bluemix (IoT)
    Research Team: York University: Brian Ramprasad  Marios Fokaefs  Marin Litoiu 
    IBM: Joe Wigglesworth 

  • Improving Garbage Collection-Time String Deduplication
    Exhibit Session: Tuesday P5
    Title: Improving Garbage Collection-Time String Deduplication
    Chair(s): Konstantin Nasartschuk
    Format: Poster
    Description: Immutable string objects in Java are the most common objects on the heap. To optimize their structure, placement on the heap, and to reduce the amount of duplicate memory they consume is an important VM optimization. String deduplication is the approach of identifying duplicate character arrays and reusing them for multiple objects. We describe improvements to the approach to increase performance and dynamically enable/disable the approach when the application structure can benefit from it.
    Industry Company: IBM
    Project Number 1019
    Product Name: IBM SDK
    Web Site: http://www.unb.ca/research/casatlantic/
    Research Team: University of New Brunswick: Konstantin Nasartschuk  Kenneth B. Kent  Stephen A. MacKay 

  • On-demand, Differential, and Incremental Interprocedural Taint Vulnerability Analysis
    Exhibit Session: Tuesday P8
    Title: On-demand, Differential, and Incremental Interprocedural Taint Vulnerability Analysis
    Chair(s): Nicolas Cloutier , Ettore Merlo
    Format: Poster
    Description: Investigate incremental approaches based on code changes to reduce the computing time of tainted traces. Make use of fixpoints on the data flow (SSA) to investigate differentials and incrementals approaches. Computation time is advantageously proportional to only the size of code changes. Execution times from experiments on WebGoat are short and promising for the future research.
    Industry Company: IBM
    Project Number 1044
    Product Name: IBM Security AppScan Source
    Research Team: Ecole Polytechnique de Montreal: Nicolas Cloutier  Ettore Merlo 
    HCL Technologies: John Peyton 
    IBM Security Systems: Babita Sharma 

  • Private vitals data as a determinant of collaborative care quality
    Exhibit Session: Wednesday P9
    Title: Private vitals data as a determinant of collaborative care quality
    Chair(s): Abidin Akkok , Peter Pennefather , Deb Fels , Patrick Neumann
    Format: Poster
    Description: A conceptual framework for sharing private health data and in particular personal vital signs data from patient-owned digital health devices is presented. It defines a unified storage strategy for all kinds of vital signs data collected from all kinds of devices with a key proviso that it is first registered and stored as patient-owned private data. A system diagram of the conceptual framework showing the relationships and information flow for monitoring care service quality is described.
    Research Team: Ryerson University: Z. Abidin Akkok,  Deborah I. Fels,  W. Patrick Neumann 
    gDial Inc.: Peter Pennefather,  West Suhanic 

  • Quantifying Duplication to Improve Data Quality
    Exhibit Session: Monday P11
    Title: Quantifying Duplication to Improve Data Quality
    Chair(s): Yu Huang
    Format: Poster
    Description: We present a record deduplication framework that differentiates terms during the matching process to improve overall accuracy. We also define a duplication metric that quantifies the level of duplication for an attribute value, and within an attribute. This metric can be used by analysts to understand the distribution and similarity of values during the data cleaning process.
    Industry Company: IBM
    Project Number 923
    Product Name: IBM Watson Analytics
    Research Team: McMaster University: Yu Huang  Fei Chiang 
    IBM Ottawa: Martin Petitclerc 
    IBM Germany: Albert Maier  Yannick Saillet 
    IBM Chicago: Damir Spisic 
    IBM Toronto: Calisto Zuzarte 

  • Run-time specialization of parametric kernels for graphics processing units
    Exhibit Session: Tuesday P12
    Title: Run-time specialization of parametric kernels for graphics processing units
    Chair(s): Marc Moreno Maza , Masoud Ataei Jaliseh , Haoze Yuan
    Format: Poster
    Description: Parametric kernels are GPU programs which are generated from annotated C/C++ code . In general, during this generation process, machine and program parameters may not be determined accurately; for this reason they are manipulated as unknown symbols until these GPU programs are actually run. This poster describes how the values of those parameters are obtained at run-time.
    Industry Company: IBM
    Project Number 880
    Product Name: IBM XL
    Web Site: www.metafork.org
    Research Team: University of Western Ontarioo: Marc Moreno-Maza  Masoud Ataei  Haoze Yuan 
    IBM Canada: Jeeva Paudel 

  • Sentimental eCrits
    Exhibit Session: Tuesday P13
    Title: Sentimental eCrits
    Chair(s): Lloyd Montgomery , Gabriel Tapuc , Tyson Bulmer , Rickus Senekal
    Format: Poster
    Description: Sentimental eCrits is the second stage of the eCrits project, a project designed to predict escalations against support tickets, particularly Critical Situations against PMRs. The Sentimental stage involves our newly acquired access to the text of PMRs, so now the sentiment of customer and support conversations is being analyzed and reported as a metric against which escalations can be predicted. The sentiment is being extracted using Watson APIs.
    Industry Company: IBM
    Project Number 1007
    Product Name: Watson-based cognitive support to enhance customer supp
    Research Team: University of Victoria: Lloyd, Gabriel, Tyson, Rickus 

  • ShopBuddy: Store Path Optimization
    Exhibit Session: Tuesday P14
    Title: ShopBuddy: Store Path Optimization
    Chair(s): Prashanti Angara , Athabasca Witschi , Ulrike Stege , Hausi Muller
    Format: Poster
    Description: The ShopBuddy application aims to suggest paths through stores to customers so that they can shop efficiently while taking into consideration the store's aims at maximizing profits. We present research on a few problem formulations in this regard and how variants of the Traveling Salesman Problem may be used to suggest paths to shoppers.
    Research Team: University Of Victoria: Prashanti Angara  Athabasca Witschi  Ulrike Stege  Hausi Mller 

  • Taming services on the Bluemix
    Exhibit Session: Tuesday P15
    Title: Taming services on the Bluemix
    Chair(s): Naresh Eeda , Nikita Sokolov
    Format: Poster
    Description: This research focuses on: i. Migrating the Bluemix Dashboard to a GraphQL-based UI Architecture to achieve key performance improvements. ii. Creating a smart monitoring system for Bluemix UI, capable of learning and identifying abnormal behavioral patterns with the use of unsupervised and semi-supervised learning algorithms. iii. Alleviating deployment complexity by adopting the use of a container management system, such as kubernetes.
    Industry Company: IBM
    Project Number 1045
    Product Name: Taming services on the Bluemix
    Research Team: The University of Western Ontario: Naresh Eeda, Nikita Sokolov, Jehad Fares 

  • The DevOps/BizOps Integration: A New Vision for Software Engineering
    Exhibit Session: Wednesday P14
    Title: The DevOps/BizOps Integration: A New Vision for Software Engineering
    Chair(s): Marios Fokaefs
    Format: Poster
    Description: Following new advancements in software, DevOps has risen as a new development paradigm, one that promotes the integration between development and management even after the development and deployment of the system. As digital businesses are also on the rise and software has become ubiquitous as a product or as a tool, we also argue the need for integration with BizOps. We describe an integrated and hierarchical decision process combining technical, economic and business parameters.
    Research Team: Polytechnique Montral: Marios Fokaefs 
    York University: Marin Litoiu 

  • Tuning Spark Performance Aiming for Various Types of Workloads
    Exhibit Session: Wednesday P15
    Title: Tuning Spark Performance Aiming for Various Types of Workloads
    Chair(s): Jenne Zhang
    Format: Poster
    Description: As a unified engine for big data analytics, Spark provides high scalability and fault-tolerance with its unique in-memory engine. Ensuring its ease of use, Spark hides the complex settings from users, however, there are approximately 200 configurable parameters in Spark execution engine. The default values might not be the best setting for every workload. My research goal is to develop a tool to help users tune Spark for different types of workloads and to improve overall performance.
    Research Team: Queen's University: Jie ni Zhang, Patrick Martin 

See details below:

Exhibit Sessions - 11:45 to 12:45 and 17:00 to 19:30

  • D1: A Hierarchical and Dynamic Security Control System for Large Scale Systems
    Booth Number: D1
    Title: A Hierarchical and Dynamic Security Control System for Large Scale Systems
    Chair(s): Yar Rouf , Marin Litoiu
    Format: Demo
    Description: We propose a hierarchical distributed architecture for security control in order to partition responsibility and workload among many security controllers. Our architecture proposes a more simplified way of defining security rules to allow security to be enforced on an operational level, rather than a development level. We demonstrate the ability of our system to expand with additional sensors, traffic data and detection rules at run time in a systematic manner.
    Industry Company: IBM
    Project Number 1004
    Product Name: IBM PureApp
    Research Team: York University: Yar Rouf,   Mark Shtern,   Marios Fokaefs,   Marin Litoiu 
    IBM: Vio Onut 

  • P1: A Scalability-oriented Benchmark Suite for Node.js on the Cloud
    Booth Number: P1
    Title: A Scalability-oriented Benchmark Suite for Node.js on the Cloud
    Chair(s): Jiapeng Zhu , Panagiotis (Panos) Patros
    Format: Poster
    Description: Clouds provide on-demand and pay-as-you-go computing resources to their end users. Node.js is a popular choice for cloud applications as it is event-driven, asynchronous and non-blocking. However, Node.js underutilizes multi-core CPUs due to its single-threaded nature, which limits its scalability on the cloud. Because no scalability-oriented benchmark suite exists for Node.js, we will focus on developing one. Our benchmarking suite will facilitate and quantify Node.js scalability improvements.
    Industry Company: IBM
    Project Number 1002
    Product Name: IBM SDK for Node.js
    Web Site: http://www.unb.ca/research/casatlantic/
    Research Team: University of New Brunswick: Jiapeng Zhu, Patros Panagiotis, Kenneth Kent 
    IBM Canada: Michael Dawson 

  • P2: A Scalable Architecture for the Internet of Things
    Booth Number: P2
    Title: A Scalable Architecture for the Internet of Things
    Chair(s): Brian Ramprasad , Marios Fokaefs , Marin Litoiu
    Format: Poster
    Description: As the volume and variety of sensors continues to grow, a solution is needed to process large volumes of streaming data produced by IoT devices. We propose an architecture that is flexible and scalable and an emulation framework to test the scalability of large scale sensor networks. Our design employs IoT gateways using Node-RED, a messaging broker as an aggregation point, and a streaming engine to pre-process and persist the IoT data.
    Industry Company: IBM
    Project Number 1004
    Product Name: IBM Bluemix (IoT)
    Research Team: York University: Brian Ramprasad  Marios Fokaefs  Marin Litoiu 
    IBM: Joe Wigglesworth 

  • D2: Accessibility in the DevOps Era
    Booth Number: D2
    Title: Accessibility in the DevOps Era
    Chair(s): Devan Shah , Larry Lindsay , ali asghar
    Format: Demo
    Description: As enterprises adopt continuous deployment practices, the DevOps testing practice must adapt accessibility testing tools and processes or user experience will suffer. These tools must work hand in hand with your application DevOps methodology - both required by & enabling those practices. In this hands-on demo, the audience will learn about IBMs cutting-edge accessibility DevOps tooling and see how simple it is to integrate the tools into an existing continuous integration pipeline (CI).
    Web Site: https://ibm.biz/a11yDashboard
    Research Team: IBM: Devan Shah  Larry Lindsay  Thomas Brunet  Ali Asghar 

  • P3: An Automatic Approach for Transforming IoT Applications to RESTful Services
    Booth Number: P3
    Title: An Automatic Approach for Transforming IoT Applications to RESTful Services
    Chair(s): Yu Zhao , Yongjian Yang
    Format: Poster
    Description: Internet of Things (IoT) devices are prevalent in all aspects of our lives. Nowadays, IoT devices are controlled by various end-user applications. In this context, the functionalities of IoT devices may be published as IoT services. The uniform interface of IoT services allows them to be integrated with existing applications. We propose an approach that automatically transforms functionalities of IoT devices to IoT services hosted on the cloud.
    Industry Company: IBM
    Project Number 920
    Product Name: Service Composition
    Research Team: Queen's University: Yu Zhao, Yongjian Yang, Ying Zou 
    IBM: Joanna Ng 

  • D3: Classification Algorithms and How to Distribute Them
    Booth Number: D3
    Title: Classification Algorithms and How to Distribute Them
    Chair(s): Rebecca Young
    Format: Demo
    Description: This session discusses a Distributed Classifier Training (DCT) approach as a free open-source IBM SPSS Modeler plugin to address the current challenges of adapting classification algorithms in handling large volumes of data from big data environments.
    Industry Company: IBM
    Business unit Analytics
    Product Name: IBM SPSS Modeler
    Research Team: IBM: Rebecca Young 

  • D4: DevOps Infrastructures to Support Continuous Software Deployment
    Booth Number: D4
    Title: DevOps Infrastructures to Support Continuous Software Deployment
    Chair(s): Kostas Kontogiannis
    Format: Demo
    Description: An approach that is attracting the attention of the software engineering community towards minimizing deployment risks while streamlining the delivery of new product features on a continual basis without sacrificing quality is release small release often. In order to support such an approach efficient DevOps pipelines have to be assembled. Such pipelines involve tool-chains and infrastructure that uses software analytics and decision support systems to reach release/no-release decisions.
    Industry Company: IBM
    Project Number 1048
    Product Name: IBM DevOps Insights
    Web Site: www.csd.uwo.ca/~kostas
    Research Team: Western University: Marios Grigoriou: Ph.D. Candidate, Western U  Kostas Kontogiannis: Professor, Western Univ  Don Cronin: Program Director, Lead Architect  Alberto Giammaria: STSM, IBM Watson & Cloud   DevOps Analytics  Chris Brealey: STSM - IBM STSM - IBM Cloud,  

  • P4: Hybrid Scaling of Dockerized Microservices Architectures in Cloud Data Centres
    Booth Number: P4
    Title: Hybrid Scaling of Dockerized Microservices Architectures in Cloud Data Centres
    Chair(s): Anthony Kwan
    Format: Poster
    Description: Our research presents an experimental platform that enables the evaluation of various resource adjustment algorithms. Two novel hybrid autoscaling algorithms are presented and benchmarked against the Kubernetes horizontal scaling algorithm.
    Industry Company: IBM
    Project Number 1008
    Product Name: IBM Blueworks Live
    Research Team: University of Toronto: Anthony Kwan  Jonathon Wong  Hans-Arno Jacobsen 
    IBM: Vinod Muthuswamy  Suzette Samoojh  Allen Chan 

  • P5: Improving Garbage Collection-Time String Deduplication
    Booth Number: P5
    Title: Improving Garbage Collection-Time String Deduplication
    Chair(s): Konstantin Nasartschuk
    Format: Poster
    Description: Immutable string objects in Java are the most common objects on the heap. To optimize their structure, placement on the heap, and to reduce the amount of duplicate memory they consume is an important VM optimization. String deduplication is the approach of identifying duplicate character arrays and reusing them for multiple objects. We describe improvements to the approach to increase performance and dynamically enable/disable the approach when the application structure can benefit from it.
    Industry Company: IBM
    Project Number 1019
    Product Name: IBM SDK
    Web Site: http://www.unb.ca/research/casatlantic/
    Research Team: University of New Brunswick: Konstantin Nasartschuk  Kenneth B. Kent  Stephen A. MacKay 

  • P6: Modernizing Inlining Strategies in Just-in-Time Compilers
    Booth Number: P6
    Title: Modernizing Inlining Strategies in Just-in-Time Compilers
    Chair(s): Erick Ochoa Lopez
    Format: Poster
    Description: Method inline substitution is traditionally driven by a greedy algorithm that minimizes the number of dynamic method calls. Instead of minimizing the number of dynamic method calls, we propose to guide inlining decisions by the optimizations that are enabled after inlining has happened.
    Industry Company: IBM
    Project Number 1047
    Product Name: IBM Java SDK
    Research Team: University of Alberta: Erick Ochoa Lopez  Karim Ali  J. Nelson Amaral 
    IBM Canada: Andrew Craik 

  • P7: OMR: A robust open-source language runtime toolkit
    Booth Number: P7
    Title: OMR: A robust open-source language runtime toolkit
    Chair(s): Xiaoli Liang , Arianne Butler , Daryl Maier
    Format: Poster
    Description: The Eclipse OMR project is a set of open-source C and C++ components that can be used to build robust language runtimes that support many different hardware and operating system platforms. By using the Eclipse OMR project, language implementers can create full featured languages more quickly and easily to enrich the options available to programmers. We have implemented working Ruby and Lua prototypes using OMR, in addition to OpenJ9 which will soon be released.
    Industry Company: IBM
    Business unit IBM Cloud
    Product Name: OMR
    Web Site: https://github.com/eclipse/omr/
    Research Team: IBM Canada: Arianne Butler  Daryl Maier  Xiaoli Liang 

  • P8: On-demand, Differential, and Incremental Interprocedural Taint Vulnerability Analysis
    Booth Number: P8
    Title: On-demand, Differential, and Incremental Interprocedural Taint Vulnerability Analysis
    Chair(s): Nicolas Cloutier , Ettore Merlo
    Format: Poster
    Description: Investigate incremental approaches based on code changes to reduce the computing time of tainted traces. Make use of fixpoints on the data flow (SSA) to investigate differentials and incrementals approaches. Computation time is advantageously proportional to only the size of code changes. Execution times from experiments on WebGoat are short and promising for the future research.
    Industry Company: IBM
    Project Number 1044
    Product Name: IBM Security AppScan Source
    Research Team: Ecole Polytechnique de Montreal: Nicolas Cloutier  Ettore Merlo 
    HCL Technologies: John Peyton 
    IBM Security Systems: Babita Sharma 

  • P10: Optimizing the JVM Object Model Using Object Splitting
    Booth Number: P10
    Title: Optimizing the JVM Object Model Using Object Splitting
    Chair(s): Taees Eimouri , Kenneth Kent , Aleksandar Micic
    Format: Poster
    Description: Data layout optimization is a well-known method to improve cache performance by reorganizing data elements. We introduce a novel approach to optimize layout of objects, called the Object Splitting Technique, whereby Java objects are split at allocation time so that those fields of the split objects that are not accessed as often are separated from the rest of the fields. We implemented the approach in IBM's JVM. Tests with different benchmarks, in most cases, reduced the number of cache misses.
    Industry Company: IBM
    Project Number 1021
    Product Name: IBM SDK
    Web Site: http://www.unb.ca/research/casatlantic/
    Research Team: University of New Brunswick: Taees Eimouri  Dr. Kenneth B. Kent 
    IBM Canada: Aleksandar Micic 

  • P9: OptImatch: Semantic-Web System for Automated Discovery of Query Problem Patterns
    Booth Number: P9
    Title: OptImatch: Semantic-Web System for Automated Discovery of Query Problem Patterns
    Chair(s): Guilherme Damasio
    Format: Poster
    Description: OptImatch is a graph-based system that offers a way to look for varied user defined problem patterns in Query Execution Plans (QEPs) and automatically get recommendations from an expert provided and user customizable knowledge base. As current work, we are extending the tool to automatically discover, based on previously knowledge, the best QEP to be applied to a given query and to automatically give recommendations such as changing database configuration in order to improve the performance.
    Industry Company: IBM
    Project Number 1049
    Product Name: OptImatch
    Research Team: UOIT: Guilherme Damasio,  Jarek Szlichta 
    York University: Parke Godfrey 
    IBM: Calisto Zuzarte,  Piotr Mierzejewski,  Vincent Corvinelli 

  • P11: Post-mortem debugging with Promises for Node.js
    Booth Number: P11
    Title: Post-mortem debugging with Promises for Node.js
    Chair(s): Maxim Uzun , Patros Panagiotis
    Format: Poster
    Description: As larger distributed systems increase their reliance on Node.js, tools to better understand complex software failures become more important. Post-mortem debugging for dynamic environments is behind that of native environments. Given the asynchronous nature of Node.js, a function passed to a Promise object is executed on the next tick and any unhandled rejections will show at a later time. We are investigating improved debugging methods for Node.js when using Promises and evaluating their cost.
    Industry Company: IBM
    Project Number 1002
    Product Name: IBM SDK for Node.js
    Web Site: http://www.unb.ca/research/casatlantic/
    Research Team: University of New Brunswick: Maxim Uzun, Patros Panagiotis, Kenneth Kent 
    IBM Canada: Michael Dawson 

  • P12: Run-time specialization of parametric kernels for graphics processing units
    Booth Number: P12
    Title: Run-time specialization of parametric kernels for graphics processing units
    Chair(s): Marc Moreno Maza , Masoud Ataei Jaliseh , Haoze Yuan
    Format: Poster
    Description: Parametric kernels are GPU programs which are generated from annotated C/C++ code . In general, during this generation process, machine and program parameters may not be determined accurately; for this reason they are manipulated as unknown symbols until these GPU programs are actually run. This poster describes how the values of those parameters are obtained at run-time.
    Industry Company: IBM
    Project Number 880
    Product Name: IBM XL
    Web Site: www.metafork.org
    Research Team: University of Western Ontarioo: Marc Moreno-Maza  Masoud Ataei  Haoze Yuan 
    IBM Canada: Jeeva Paudel 

  • D5: Semantic Aware Anomaly Detection on the Cloud
    Booth Number: D5
    Title: Semantic Aware Anomaly Detection on the Cloud
    Chair(s): Arnamoy Bhattacharyya , Stelios Sotiriadis
    Format: Demo
    Description: We introduce a new tool for phase detection and characterization for applications running on the cloud. We evaluate our tool for a number of C, C++ and Java application servers. Our tool is integrated with an IBM Watson Conversation engine so that administrators can know about system health using simple texts.
    Industry Company: IBM
    Project Number 916
    Product Name: Bluemix
    Research Team: University of Toronto: Arnamoy, Stelios, Cristiana Amza 
    IBM Toronto: Joe Wigglesworth 

  • P13: Sentimental eCrits
    Booth Number: P13
    Title: Sentimental eCrits
    Chair(s): Lloyd Montgomery , Gabriel Tapuc , Tyson Bulmer , Rickus Senekal
    Format: Poster
    Description: Sentimental eCrits is the second stage of the eCrits project, a project designed to predict escalations against support tickets, particularly Critical Situations against PMRs. The Sentimental stage involves our newly acquired access to the text of PMRs, so now the sentiment of customer and support conversations is being analyzed and reported as a metric against which escalations can be predicted. The sentiment is being extracted using Watson APIs.
    Industry Company: IBM
    Project Number 1007
    Product Name: Watson-based cognitive support to enhance customer supp
    Research Team: University of Victoria: Lloyd, Gabriel, Tyson, Rickus 

  • P14: ShopBuddy: Store Path Optimization
    Booth Number: P14
    Title: ShopBuddy: Store Path Optimization
    Chair(s): Prashanti Angara , Athabasca Witschi , Ulrike Stege , Hausi Muller
    Format: Poster
    Description: The ShopBuddy application aims to suggest paths through stores to customers so that they can shop efficiently while taking into consideration the store's aims at maximizing profits. We present research on a few problem formulations in this regard and how variants of the Traveling Salesman Problem may be used to suggest paths to shoppers.
    Research Team: University Of Victoria: Prashanti Angara  Athabasca Witschi  Ulrike Stege  Hausi Mller 

  • P15: Taming services on the Bluemix
    Booth Number: P15
    Title: Taming services on the Bluemix
    Chair(s): Naresh Eeda , Nikita Sokolov
    Format: Poster
    Description: This research focuses on: i. Migrating the Bluemix Dashboard to a GraphQL-based UI Architecture to achieve key performance improvements. ii. Creating a smart monitoring system for Bluemix UI, capable of learning and identifying abnormal behavioral patterns with the use of unsupervised and semi-supervised learning algorithms. iii. Alleviating deployment complexity by adopting the use of a container management system, such as kubernetes.
    Industry Company: IBM
    Project Number 1045
    Product Name: Taming services on the Bluemix
    Research Team: The University of Western Ontario: Naresh Eeda, Nikita Sokolov, Jehad Fares 

  • D6: Using IBM Watson cloud services to analyze chat conversations and forum posts
    Booth Number: D6
    Title: Using IBM Watson cloud services to analyze chat conversations and forum posts
    Chair(s): Sarah Packowski
    Format: Demo
    Description: Chat tools and forums are popular for engaging with potential customers and supporting users. Beyond their effectiveness at answering questions in the moment, these tools provide valuable insight into what potential customers and users are thinking, saying, and doing. This exhibit demonstrates how to use IBM Watson cloud services to identify trends, pain points, and customer satisfaction in chat conversations and forums.
    Business unit IBM Hybrid Cloud
    Product Name: IBM Watson cloud services
    Web Site: http://ibm.biz/CASCON-2017_Analyzing_chat
    Research Team: IBM: Sarah Packowski 

  • D7: Using the IBM OpenMP XL compiler to accelerate scientific applications
    Booth Number: D7
    Title: Using the IBM OpenMP XL compiler to accelerate scientific applications
    Chair(s): Ettore Tiotto , Kelvin Li , Whitney Tsang , Bardia Mahjour
    Format: Demo
    Description: The IBM XL compilers are used to accelerate scientific computation on IBM POWER system connected NVIDIA Pascal GPUs. The compilers supports the new OpenMP 4.5 programming model which allows offloading computation to GPUs easily and with excellent performance, and is much easier to use than the CUDA programming model.
    Product Name: IBM XL C/C++ and Fortran compilers
    Research Team: IBM Canada Laboratory: Ettore Tiotto  Bardia Mahjour  Kelvin Li  Whitney Tsang  Wael Yehia  Samer Jacob   Tarique Islam  Xing Xue  Robert Ho  Jeeva Paudel 

  • P16: XL Compiler Extensions for GPU Platforms
    Booth Number: P16
    Title: XL Compiler Extensions for GPU Platforms
    Chair(s): Akihiro Hayashi , Jun Shirako
    Format: Poster
    Description: In this CAS project, we explore compile-time optimizations for OpenMP accelerator programs targeting the POWER8+GPUs platform. We build on existing optimizations in the TPO component of the IBM XL compiler, by introducing: aggressive dependence analysis of explicitly parallelized OpenMP programs; and loop transformations to enhance GPU parallelism and memory efficiency.
    Industry Company: IBM
    Project Number 896
    Product Name: IBM XL C/C++ compiler, XL Fortran compiler
    Research Team: Rice University: Jun Shirako  Akihiro Hayashi  Vivek Sarkar 
    IBM Canada: Robert Ho  Ettore Tiotto 


Expo Reception / Short Talks / PechaKucha Expo Presentations

Welcome - 17:15 to 17:30

Short Talk Sessions - 17:30 to 18:15 (3 talks).   Talks order as follows (15 mins each)

  • S1: Making Stuff Fast
    Booth Number: S1
    Title: Making Stuff Fast
    Chair(s): Mark Stoodley
    Format: Short Talk
    Description: It's never been easier to create your own Just In Time Compiler, be it for a language, for an application, or just to learn how! Learn how to use the JitBuilder library at the open source Eclipse OMR project to dynamically generate executable native code that does whatever you want! This library has been used to dynamically compile Swift, Javascript, Lua, Smalltalk, the Rosie Pattern Language, the LLVM Kaleiedoscope tutorial language, and an educational runtime called Base9. What will you build?
    Product Name: IBM Canada
    Web Site: www.github.com/eclipse/omr
    Research Team: IBM Canada: Mark 

  • S2: Phishing clustering based on MST
    Booth Number: S2
    Title: Phishing clustering based on MST
    Chair(s): Qian Cui
    Format: Short Talk
    Description: We provide a new method to analyze and detect phishing attacks based on minimal spanning tree (MST). Compared with other models, this new method is able to track more effectively the evolution of phishing attacks and the similarities between related instances. Our method creates clusters of related attack instances based on an MST of the cluster and a locally defined threshold. Our method is able to detect more than 90% of the new attack instances in our experiments.
    Research Team: University of Ottawa: Qian Cui, Ph.D. candidate 

  • S3: Supporting Microservice Evolution
    Booth Number: S3
    Title: Supporting Microservice Evolution
    Chair(s): Julia Rubin
    Format: Short Talk
    Description: Microservices have become a popular pattern for deploying scale-out application logic. An advantage of using microservices is their loose coupling, which leads to agile and rapid evolution, and continuous re-deployment. However, developers are tasked with managing this evolution and largely do so manually by continuously evaluating low-level service behaviors. This is tedious, error-prone, and slow. In this talk, we discuss how our approach can help engineers with these tasks.
    Industry Company: IBM
    Project Number 1040
    Product Name: IBM Bluemix
    Research Team: University of British Columbia: Harshavardhan Kadiyala,   Bo Hu,  Ivan Beschastnikh,   Julia Rubin 
    IBM: John Steinbacher,   Tony Erwin, 
    Federal University of Pernambuco, Brazil: Adalberto R. Sampaio Jr.  Nelson Rosa 

Break - 18:15 to 18:25

PechaKucha Sessions - 18:25 to 19:30. (13 talks).   Presentations order as follows (5 mins each)

  • Center for Health Informatics and Analytics
    Exhibit Session: Monday P4
    Title: Center for Health Informatics and Analytics
    Chair(s): Randy Giffen
    Format: Poster
    Description: The Center for Health Informatics and Analytics is a collaboration between Memorial University and IBM Canada. The goal of the center is to jointly develop and validate solutions that help translate knowledge and data into action. The solutions will support both clinical and administrative scenarios. The center is based on a vision for healthcare transformation. Lower healthcare costs can be achieved by improving efficiency and effectiveness and ultimately by improving health.
    Industry Company: IBM
    Product Name: Center for Health Informatics and Analytics
    Web Site: http://www.med.mun.ca/CHIA/Home.aspx
    Research Team: IBM Canada: Randy Giffen 

  • Cognitive Telescope Network
    Exhibit Session: Wednesday P1
    Title: Cognitive Telescope Network
    Chair(s): Arunava Majumdar
    Format: Poster
    Description: Telescopic follow-up of transient astronomical events is one of the most desirable and scientifically useful activities in modern observational astronomy. Pinpointing a transient is essential for discovering more about the source by directing more powerful telescopes to observe and maintain records as the transient evolves. Join us to see how IBM Watson Visual Recognition, Machine Learning and Conversation are used to help build the telescope network on Bluemix.
    Business unit Cloud
    Product Name: Cognitive Telescope Network
    Research Team: IBM: Arunava Majumdar 

  • Cold Object Segregation
    Exhibit Session: Monday P5
    Title: Cold Object Segregation
    Chair(s): Scott Young , Kenneth Kent , Gerhard Dueck
    Format: Poster
    Description: Cold (infrequently referenced) objects can take up extra space in the areas of memory reserved for frequently accessed, tenured objects. Identifying all cold objects in tenured space during runtime is expensive. This project seeks to find an algorithm for selecting candidate cold objects to be moved to secondary storage so that the difference between time saved during garbage collection and time spent performing the selection is maximized.
    Industry Company: IBM
    Project Number 1001
    Product Name: OMR and J9
    Web Site: http://www.unb.ca/research/casatlantic/
    Research Team: UNB: Scott Young  Kenneth Kent  Gerhard Dueck 
    IBM Canadda: Charlie Gracie 

  • MASS library scalar math functions for the Go Language
    Exhibit Session: Monday P9
    Title: MASS library scalar math functions for the Go Language
    Chair(s): Bill O'Farrell
    Format: Poster
    Description: Improving scalar math performance in the Go programming language on IBM Z. We took the scalar MASS libraries and translated them from Z assembly to Go assembly, had them approved as open source, and contributed them to Go. The translation was done automatically (mostly) via a python script. The project is complete and was included in go 1.9.
    Industry Company: IBM
    Business unit Systems
    Product Name: Go Language (Open-Source) on IBM Z
    Research Team: IBM: Bill O'Farrell  Kang Zhang  Annita Zhang  Robert Enenkel 

  • NUMA GC
    Exhibit Session: Monday P6
    Title: NUMA GC
    Chair(s): Maria Patrou
    Format: Poster
    Description: Non-uniform Memory Access (NUMA) systems use different types of memory accesses. IBM's Java Virtual Machine identifies a NUMA architecture, while using memory and threads from the available nodes in a distributed way. A design for a node-isolated memory achieved by a node-heap resize functionality and a thread policy is proposed. Finally, different modes regarding hardware and thread characteristics are investigated to identify the application attributes that can benefit from specific modes.
    Industry Company: IBM
    Project Number 1020
    Product Name: IBM SDK
    Web Site: http://www.unb.ca/research/casatlantic/
    Research Team: University of New Brunswick: Maria Patrou  Kenneth B. Kent  Gerhard W. Dueck 
    IBM Canada: Charlie Gracie  Aleksandar Micic 

  • OptImatch: Semantic-Web System for Automated Discovery of Query Problem Patterns
    Exhibit Session: Tuesday P9
    Title: OptImatch: Semantic-Web System for Automated Discovery of Query Problem Patterns
    Chair(s): Guilherme Damasio
    Format: Poster
    Description: OptImatch is a graph-based system that offers a way to look for varied user defined problem patterns in Query Execution Plans (QEPs) and automatically get recommendations from an expert provided and user customizable knowledge base. As current work, we are extending the tool to automatically discover, based on previously knowledge, the best QEP to be applied to a given query and to automatically give recommendations such as changing database configuration in order to improve the performance.
    Industry Company: IBM
    Project Number 1049
    Product Name: OptImatch
    Research Team: UOIT: Guilherme Damasio,  Jarek Szlichta 
    York University: Parke Godfrey 
    IBM: Calisto Zuzarte,  Piotr Mierzejewski,  Vincent Corvinelli 

  • PANORAMA: Deployment Specification in the Context of Large Scale Systems
    Exhibit Session: Monday D7
    Title: PANORAMA: Deployment Specification in the Context of Large Scale Systems
    Chair(s): Miguel Jimenez , Hausi Muller , Gabriel Tamura , Norha M. Villegas
    Format: Demo
    Description: DevOps has dramatically influenced how/when deployment is realised. However, little effort has been devoted to the need for deployment and configuration specifications to support the various levels of detail and abstraction present in large-scale systems. In this exhibit, we present PANORAMA, our envisioned environment to design and visualise large-scale continuous deployment based on the OASIS TOSCA specification.
    Research Team: University of Victoria: Miguel Jimnez, Hausi Mller 
    Universidad Icesi: Gabriel Tamura, Norha Villegas 

  • Phishing Numbers
    Exhibit Session: Monday P7
    Title: Phishing Numbers
    Chair(s): Guy Jourdan , Qian Cui
    Format: Poster
    Description: Are popular websites hacked to host phishing attacks? Why do phishing sites sometimes redirect their victims to other sites? How often are phishing attacks hosted on compromised servers? What are the most common paths used in phishing attacks?... We have been gathering sample phishing attacks for the past 18 months, and have conducted a number of analyses to answer some of these questions. This exhibit will provide an overview of our findings to date.
    Industry Company: IBM
    Project Number 921
    Product Name: Security Forensics
    Research Team: University of Ottawa: Guy-Vincent Jourdan  Gregor Bochmann  Qian Cui  Sophie LePage 
    IBM: Vio Onut  Russ Couturier 

  • Pressure Sensitive Mats in the NICU
    Exhibit Session: Wednesday P7
    Title: Pressure Sensitive Mats in the NICU
    Chair(s): Shermeen Nizami
    Format: Poster
    Description: To examine the potential for incorporating Pressure Sensitive Mats (PSM) and video data analysis as part of an integrated patient monitoring environment. Data will be collected simultaneously from PSM, video, and existing patient monitors, and shall be integrated with gold-standard bed-side annotations collected on an mHealth app developed as part of this project. Data analytics are being developed and evaluated to detect various events of clinical interest in the neonatal intensive care unit.
    Industry Company: IBM
    Project Number SRG_2015_1
    Product Name: Watson Analytics, Watson Health
    Research Team: Carleton University: Dr.Shermeen Nizami  Amente Bekele  Yasmina Souley Dosso  Mohamed Hozayen  Dr.James R.Green 
    IBM CAS: Dr.Randy Giffen  Nathalie Le Prohon 
    Children's Hospital of Eastern Ontario: Dr.JoAnn Harrold   Kim Greenwood 

  • Prioritize and Visualize Vulnerability Findings
    Exhibit Session: Wednesday P8
    Title: Prioritize and Visualize Vulnerability Findings
    Chair(s): Jinqiu Yang , Lin Tan
    Format: Poster
    Description: We propose to prioritize and visualize vulnerability findings of AppScan. The visualization helps developers prioritize workloads accordingly by grouping vulnerability findings with the same potential fix location. In addition, we propose a tree-based visualization to link two relevant vulnerability findings. For example, it is beneficial to link second-order SQL injections and their relevant SQL injections for easier understanding.
    Industry Company: IBM
    Project Number 911
    Product Name: AppScan
    Research Team: University of Waterloo: Jinqiu Yang  Lin Tan 
    AppScan: John Peyton  Kristofer A Duer 

  • Recommender System for IBM Watson Analytics-Algorithm Development
    Exhibit Session: Monday P12
    Title: Recommender System for IBM Watson Analytics-Algorithm Development
    Chair(s): Parisa Lak
    Format: Poster
    Description: IBM Watson Analytics (WA) provides users with visualization recommendations for their provided dataset using a rule based system. Recommender systems (RS) are a set of learning algorithms that provide personalized recommendations to the users based on their historical preferences. In this work, we design a RS that receives users historical interaction with WA as input, and provides a list of recommendations that are uniquely generated based on his/her prior interaction with the system.
    Industry Company: IBM
    Project Number 919
    Product Name: Watson Analytics
    Research Team: Ryerson University: Parisa Lak  Can Kavaklioglu  Ayse Bener  Andriy Miranskyy 
    IBM: Martin Petitclerc  Graham Wills 

  • Supporting Microservice Evolution
    Exhibit Session: Monday P14
    Title: Supporting Microservice Evolution
    Chair(s): Julia Rubin
    Format: Poster
    Description: Microservices have become a popular pattern for deploying scale-out application logic. An advantage of using microservices is their loose coupling, which leads to agile and rapid evolution, and continuous re-deployment. However, developers are tasked with managing this evolution and largely do so manually by continuously evaluating low-level service behaviors. This is tedious, error-prone, and slow. In this exhibit, we discuss how our approach can help engineers with these tasks.
    Industry Company: IBM
    Project Number 1040
    Product Name: IBM Bluemix
    Research Team: University of British Columbia: Harshavardhan Kadiyala,   Bo Hu,  Ivan Beschastnikh,   Julia Rubin 
    IBM: John Steinbacher,   Tony Erwin  
    Federal University of Pernambuco: Adalberto R. Sampaio Jr.,  Nelson Rosa 

  • Test Case Traceability
    Exhibit Session: Wednesday P13
    Title: Test Case Traceability
    Chair(s): Sravya Polisetty , Shirin Akbarinasaji
    Format: Poster
    Description: As products evolve and teams expand in size and across different locations, the link between test cases and source code files becomes weak or is sometimes lost.To enhance semantic traceability between these artifacts we use a popular trace network that uses deep learning to predict the probability score between a test case and source code file. We also compare the performance of our network with the traditional bag-of-words approach and popular tracing methods like VSM and LSI.
    Industry Company: IBM
    Project Number SRG_2014_1
    Product Name: RTC
    Research Team: Ryerson University, Toronto, Canada: Sravya Polisetty, Shirin Akbarinasaji,   Dr. Andriy Miranskyy, Dr.Ayse Bener 
    IBM, Canada: Adam Neal 

See details below:

Exhibit Sessions - 11:45 to 12:45 and 14:15 to 15:30

  • D1: Arvin II: Towards Biomechanical Gait Analytics on the Cloud
    Booth Number: D1
    Title: Arvin II: Towards Biomechanical Gait Analytics on the Cloud
    Chair(s): Sana Oladi , Panagiotis (Panos) Patros
    Format: Demo
    Description: Data analytics systems, such as IBM Watson Health Cloud, provide a secure platform for researchers and practitioners; however, they currently neglect gait biomechanics. We present Arvin II, a prototype tool, which we intend to incorporate into an analytics cloud platform. Arvin II targets force distribution data acquired from cheap, portable and even wearable devices. Such analytics is vital for neurological conditions; back pain prevention and treatment; insole design; fall prevention; etc.
    Product Name: Arvin II
    Research Team: University of New Brunswick: Sana Oladi  Panagiotis (Panos) Patros 

  • D2: Cognitive Agents with Bluemix
    Booth Number: D2
    Title: Cognitive Agents with Bluemix
    Chair(s): Adam Di Prospero , Marin Litoiu
    Format: Demo
    Description: The exhibit will demo a cognitive agent developed with IBM Bluemix technologies. The bot is primarily designed for students with disabilities looking to learn about the resources, tools and services available on the York University campus. The bot itself utilizes a number of IBM services such as Watson Conversation, Speech-to-text and Text-to-speech.
    Industry Company: IBM
    Project Number 1004
    Product Name: IBM Bluemix
    Research Team: York University: Adam Di Prospero, Marios Fokaefs,   Nojan Norouzi, Marin Litoiu  
    IBM: Joe Wigglesworth 

  • P1: Cognitive Telescope Network
    Booth Number: P1
    Title: Cognitive Telescope Network
    Chair(s): Arunava Majumdar
    Format: Poster
    Description: Telescopic follow-up of transient astronomical events is one of the most desirable and scientifically useful activities in modern observational astronomy. Pinpointing a transient is essential for discovering more about the source by directing more powerful telescopes to observe and maintain records as the transient evolves. Join us to see how IBM Watson Visual Recognition, Machine Learning and Conversation are used to help build the telescope network on Bluemix.
    Business unit Cloud
    Product Name: Cognitive Telescope Network
    Research Team: IBM: Arunava Majumdar 

  • D3: Customer Value Analysis
    Booth Number: D3
    Title: Customer Value Analysis
    Chair(s): Damir Spisic
    Format: Demo
    Description: Analytic solution provides insights on how RFM analysis can help discover the most valuable customers. Sales teams can take action based on Recency (how recent is the sale), Frequency (active customers) and Monetary (who spends the most). The solution is based on IBM Watson Analytics storybook enhanced by an application hosted on IBM Bluemix. It uses predictive modeling methods and a rich set of visualizations as well as annotations to help interpret the results.
    Industry Company: IBM
    Business unit IBM Business Analytics
    Product Name: IBM Watson Analytics
    Web Site: https://www.ibm.com/watson-analytics
    Research Team: IBM Hybrid Cloud: Damir Spisic  Keith Kroeger  Ruth Briones  Richard Oswald 

  • P2: Designing User Engagement for Cognitively-Enhanced Processes
    Booth Number: P2
    Title: Designing User Engagement for Cognitively-Enhanced Processes
    Chair(s): Zia Babar , Alexei Lapouchnian , Eric Yu
    Format: Poster
    Description: Enterprise business processes can be enhanced using cognitive capabilities by offering analytics-based decision recommendations and increased sophisticated automation in order to achieve enterprise functional objectives. Conceptual modeling techniques are proposed which allow structured analysis of these business processes as a process architecture, to guide systematic search for viable modes of interaction and cooperation between human user and cognitive advisor by studying adoption obstacles.
    Industry Company: IBM
    Project Number 1030
    Product Name: IBM BPM, IBM Blueworks
    Research Team: University of Toronto: Zia Babar, Alexei Lapouchnian, Eric Yu 
    IBM Canada: Allen Chan,   John Mourra,   Sebastian Carbajale,  Paul Pacholski,   Kam To Ngo 

  • P3: Developing serverless event driven solutions with OpenWhisk
    Booth Number: P3
    Title: Developing serverless event driven solutions with OpenWhisk
    Chair(s): NEIL DELIMA , Murali Pattathe
    Format: Poster
    Description: This exhibit will showcase IBM's serverless platform on Bluemix called Cloud Function which is based on Apache OpenWhisk. Key serverless and OpenWhisk concepts, components and use cases will be highlighted. The exhibit will include an illustration of the creation of OpenWhisk actions, sequences, triggers, rules, packages and the creation and management of APIs that wrap actions and sequences. Code snippets and sample applications will be made available for illustration and demonstration.
    Industry Company: IBM
    Business unit Hybrid Cloud
    Product Name: IBMBluemix Cloud Functions (OpenWhisk)
    Research Team: IBM Canada Lab: NEIL DELIMA 
    IBM Canada Lab: Murali Pattathe 

  • D4: Elascale: A Comprehensive Autoscaling and Monitoring Solution
    Booth Number: D4
    Title: Elascale: A Comprehensive Autoscaling and Monitoring Solution
    Chair(s): Rajsimman Ravichandiran , Hamzeh Khazaei
    Format: Demo
    Description: Autoscalability is one of the crucial features necessary for cloud software systems nowadays. Elascale strives to adjust both micro/macroservices' resources (using it's default autoscaling engine), with respect to workload and changes in the internal state of the whole application stack. The zero dependency of Elascale to the target software system enables it to be deployed as a service for any type of cloud platform.
    Product Name: Elascale
    Research Team: University of Toronto: Rajsimman Ravichandiran  Hamzeh Khazaei  

  • P4: Hybrid Similarity based Web Services Clustering (HSWSClustering)
    Booth Number: P4
    Title: Hybrid Similarity based Web Services Clustering (HSWSClustering)
    Chair(s): Waeal Obidallah
    Format: Poster
    Description: The massive number of web services over the internet with different formats and functionalities motivate our work to develop web service discovery framework based on clustering similar web services using syntactic and semantic based similarities to minimize the search space. We propose a hybrid approach that performs non-logic similarity matching for clustering web services using Cosine, WordNet, and search engine based similarities in a multilayer framework.
    Research Team: University of Ottawa: Waeal J.Obidallah, Bijan Raahemi 

  • P5: Managing developer interruption
    Booth Number: P5
    Title: Managing developer interruption
    Chair(s): Gabrielle Perez Dias , Jeremy Bradbury
    Format: Poster
    Description: The high frequency of interruptions during cognitively-intense activities can be annoying and detrimental to deadline-drive work, such as software development. Studies show that programmers tend to spend up to 30 minutes to get back to work after a pause. As interruptions are inevitable, our research focus is on the recovery of momentum. Interruption recovery is an open problem and involves understanding the interrupted activity, the developer as well as the context of the work.
    Research Team: University of Ontario IT: Gabrielle Perez Dias  Christopher Collins  Jeremy S. Bradbury 

  • D5: Managing Software Evolution Through Semantic History Slicing
    Booth Number: D5
    Title: Managing Software Evolution Through Semantic History Slicing
    Chair(s): Yi Li
    Format: Demo
    Description: Traditional commit-based sequential organization of version histories lacks semantic structure and thus are insufficient for many development tasks that require high-level, semantic understanding of program functionality, e.g., locating features and porting hot fixes. We propose to use unit tests as identifiers for corresponding software functionalities and present a family of automated techniques which analyze the semantics of historical changes and assist developers in many practical settings.
    Web Site: http://www.cs.toronto.edu/~liyi/cslicer
    Research Team: University of Toronto: Yi Li  Marsha Chechik 
    University of British Columbia: Julia Rubin 

  • P6: Modelling Design Topics in Online Discussions
    Booth Number: P6
    Title: Modelling Design Topics in Online Discussions
    Chair(s): Calahan Janik-Jones , Arthur Dailland
    Format: Poster
    Description: We demonstrate a tool that helps developers understand what topics are discussed in software design discussions in pull requests on GitHub. Specifically, we combine previous work on classifying design points in discussions with topic modelling techniques. This allows us to visualize design-specific information in a model that connects the discussion with relevant code fragments.
    Research Team: University of Toronto: Calahan Janik-Jones 
    Universit de Montral: Arthur Dailland  Michalis Famelis 

  • P7: Pressure Sensitive Mats in the NICU
    Booth Number: P7
    Title: Pressure Sensitive Mats in the NICU
    Chair(s): Shermeen Nizami
    Format: Poster
    Description: To examine the potential for incorporating Pressure Sensitive Mats (PSM) and video data analysis as part of an integrated patient monitoring environment. Data will be collected simultaneously from PSM, video, and existing patient monitors, and shall be integrated with gold-standard bed-side annotations collected on an mHealth app developed as part of this project. Data analytics are being developed and evaluated to detect various events of clinical interest in the neonatal intensive care unit.
    Industry Company: IBM
    Project Number SRG_2015_1
    Product Name: Watson Analytics, Watson Health
    Research Team: Carleton University: Dr.Shermeen Nizami  Amente Bekele  Yasmina Souley Dosso  Mohamed Hozayen  Dr.James R.Green 
    IBM CAS: Dr.Randy Giffen  Nathalie Le Prohon 
    Children's Hospital of Eastern Ontario: Dr.JoAnn Harrold   Kim Greenwood 

  • P8: Prioritize and Visualize Vulnerability Findings
    Booth Number: P8
    Title: Prioritize and Visualize Vulnerability Findings
    Chair(s): Jinqiu Yang , Lin Tan
    Format: Poster
    Description: We propose to prioritize and visualize vulnerability findings of AppScan. The visualization helps developers prioritize workloads accordingly by grouping vulnerability findings with the same potential fix location. In addition, we propose a tree-based visualization to link two relevant vulnerability findings. For example, it is beneficial to link second-order SQL injections and their relevant SQL injections for easier understanding.
    Industry Company: IBM
    Project Number 911
    Product Name: AppScan
    Research Team: University of Waterloo: Jinqiu Yang  Lin Tan 
    AppScan: John Peyton  Kristofer A Duer 

  • P9: Private vitals data as a determinant of collaborative care quality
    Booth Number: P9
    Title: Private vitals data as a determinant of collaborative care quality
    Chair(s): Abidin Akkok , Peter Pennefather , Deb Fels , Patrick Neumann
    Format: Poster
    Description: A conceptual framework for sharing private health data and in particular personal vital signs data from patient-owned digital health devices is presented. It defines a unified storage strategy for all kinds of vital signs data collected from all kinds of devices with a key proviso that it is first registered and stored as patient-owned private data. A system diagram of the conceptual framework showing the relationships and information flow for monitoring care service quality is described.
    Research Team: Ryerson University: Z. Abidin Akkok,  Deborah I. Fels,  W. Patrick Neumann 
    gDial Inc.: Peter Pennefather,  West Suhanic 

  • P10: Profiling Billions of Triples: The Case of Freebase Data Dumps
    Booth Number: P10
    Title: Profiling Billions of Triples: The Case of Freebase Data Dumps
    Chair(s): Niel Chah
    Format: Poster
    Description: Freebase was a Semantic Web knowledge base that was acquired by Google to support its Knowledge Graph and provide structured answers in Google Search. After its shutdown in 2016, Freebase data is archived in a data dump of billions of RDF triples. In this research, we present best practices in using the rich semantic data for applications such as information retrieval, present various data analytics findings that profile the data, and propose improvements to compress the data further.
    Web Site: https://github.com/nchah/freebase-triples
    Research Team: University of Toronto: Niel Chah  Periklis Andritsos 

  • P11: SlackBot for API Connect
    Booth Number: P11
    Title: SlackBot for API Connect
    Chair(s): Jisoo Lee , Laven Sathiyanathan , david cheung , ivy ho
    Format: Poster
    Description: Automation of API Connect resources through Slack integration. With this SlackBot, a number of errands could be done, such as creating or rebuilding on-prem environment based on a code release version, checking status of on-prem environment, etc. by typing commands on specific slack channel. This will save time for all API Connect users who need on-prem environment for problem reproduction, debugging, creating course content or demos.
    Industry Company: IBM
    Business unit IBM Cloud
    Product Name: API Connect
    Web Site: https://github.com/elkorep/stack-bot
    Research Team: IBM Canada Ltd.: Jisoo Lee  David Cheng  Ivy Ho  Peter El-koreh  Amir Jafarian  Bosco Li 
    Carleton University: Laven Sathiyanathan 

  • P12: Social Media for Support - Research Overview
    Booth Number: P12
    Title: Social Media for Support - Research Overview
    Chair(s): Aindrila Basak , Mona Nashaat
    Format: Poster
    Description: The poster gives an overview of a framework for building, evaluating and validating machine learning models. The framework aims at involving the end-users in every step of the model construction pipeline.The framework includes two machine learning models, where one of the models focuses on generating interpretable predictions, the other model aims at leading end-users to successful actions for each prediction.
    Research Team: University of Alberta: Aindrila Basak 
    University of Alberta: Mona Nashaat 
    University of Alberta: Prof. Dr. James Miller 
    IBM Canada: Shaikh Quader 

  • D6: SOSCIP
    Booth Number: D6
    Title: SOSCIP
    Chair(s): Elissa Strome , Natalia Mykhaylova
    Format: Demo
    Description: Established in 2012, SOSCIP is a collaborative R&D consortium that brings together academic and industry researchers to do R&D projects that leverage HPC, data analytics, AI and cognitive computing. Comprised of 15 of Ontario's most research-intensive academic institutions and IBM Canada as the founding and lead industry partner, SOSCIP has supported more than 100 collaborative projects, worked with over 60 different SMEs and helped more than 450 HQP develop data science skills and expertise.
    Industry Company: SOSCIP
    Product Name: SOSCIP
    Web Site: www.soscip.org
    Research Team: University of Toronto: Elissa Strome  Jenn MacLean  Krista Davidson  Jillian Dempsey  Natalia Mykhaylova  Sedef Akinli Kocak  Amy Hackney  Andrew Jones 

  • P13: Test Case Traceability
    Booth Number: P13
    Title: Test Case Traceability
    Chair(s): Sravya Polisetty , Shirin Akbarinasaji
    Format: Poster
    Description: As products evolve and teams expand in size and across different locations, the link between test cases and source code files becomes weak or is sometimes lost.To enhance semantic traceability between these artifacts we use a popular trace network that uses deep learning to predict the probability score between a test case and source code file. We also compare the performance of our network with the traditional bag-of-words approach and popular tracing methods like VSM and LSI.
    Industry Company: IBM
    Project Number SRG_2014_1
    Product Name: RTC
    Research Team: Ryerson University, Toronto, Canada: Sravya Polisetty, Shirin Akbarinasaji,   Dr. Andriy Miranskyy, Dr.Ayse Bener 
    IBM, Canada: Adam Neal 

  • P14: The DevOps/BizOps Integration: A New Vision for Software Engineering
    Booth Number: P14
    Title: The DevOps/BizOps Integration: A New Vision for Software Engineering
    Chair(s): Marios Fokaefs
    Format: Poster
    Description: Following new advancements in software, DevOps has risen as a new development paradigm, one that promotes the integration between development and management even after the development and deployment of the system. As digital businesses are also on the rise and software has become ubiquitous as a product or as a tool, we also argue the need for integration with BizOps. We describe an integrated and hierarchical decision process combining technical, economic and business parameters.
    Research Team: Polytechnique Montral: Marios Fokaefs 
    York University: Marin Litoiu 

  • P15: Tuning Spark Performance Aiming for Various Types of Workloads
    Booth Number: P15
    Title: Tuning Spark Performance Aiming for Various Types of Workloads
    Chair(s): Jenne Zhang
    Format: Poster
    Description: As a unified engine for big data analytics, Spark provides high scalability and fault-tolerance with its unique in-memory engine. Ensuring its ease of use, Spark hides the complex settings from users, however, there are approximately 200 configurable parameters in Spark execution engine. The default values might not be the best setting for every workload. My research goal is to develop a tool to help users tune Spark for different types of workloads and to improve overall performance.
    Research Team: Queen's University: Jie ni Zhang, Patrick Martin 

  • D7: Using IoT and cognitive services to provide a personalized experience
    Booth Number: D7
    Title: Using IoT and cognitive services to provide a personalized experience
    Chair(s): Cesar Orozco , Heath Thomann , Kyle Johnson , David Christensen
    Format: Demo
    Description: Taking advantage of IoT and cognitive services, we can build applications that provide a more personalized experience, not only by displaying personalized content, but also by changing the look and feel of the application itself to adapt it to how people feel about certain topics. You can build applications that use this combination of technologies to create a new generation of applications, making the products offered by these applications and the applications themselves more attractive
    Product Name: Rational Application Developer
    Research Team: HCL: Cesar Orozco  Heath Thomann  Kyle Johnson  David Christensen