Our Seminars in 2016

Data-Driven Acquisition of Enterprise Architecture

Presenter(s): Joel Kocherry – UOW

Date: December 1, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: It is a well known fact that the construction of an Enterprise Architecture can be a very tedious process.A technique to automate the process would be groundbreaking. To this end,a comprehensive analysis of how Enterprise Architecture can be mined from data was done using three Architecture development methods(Architmate,Zachman,TOGAF).The main objective is to lay the foundation for a tool that can be used to improve the productivity of a modeler by providing an automated process for obtaining first cut Enterprise Architecture Models.This dissertation proposes a broad class of techniques for mining a variety of Enterprise Architecture Models from readily available Enterprise data streams. It then provides a specific proposal for mining  Actor/Role and Role Catalog matrices for TOGAF in detail. It also provides a careful empirical evaluation of this technique.

Mining task post-conditions: Automating the acquisition of process semantics

Presenter(s): Metta Santiputri – UOW

Date: October 28, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: Semantic annotation of business process model in the business process designs has been addressed in a large and growing body of work, but these annotations can be difficult and expensive to acquire. This paper presents a data-driven approach to mining and validating these annotations (and specifically context-independent semantic annotations). We leverage event objects in process execution histories which describe both activity execution events (typically represented as process events) and state update events (represented as object state transition events). We present an empirical evaluation, which suggests that the approach provides generally reliable results.

Case Management in the Age of Analytics and Data-driven Insights

Presenter(s): Prof. Boualem Benatallah – Scientia Professor, School of Computer Science and Engineering, UNSW

Date: October 10, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: As economies undergo significant structural change, digital strategies and innovation must provide industries across the spectrum with tools to create a competitive edge and build more value into their services. With the advent of widely available data capture and management technologies, coupled with intensifying global competition, fluid business and social requirements, organizations are rapidly shifting to data-fication of their processes. Accordingly, they are embracing the radical changes necessary for increased productivity, added value and insights.
However, while advances in big data analytics enable tremendous automation and scalability opportunities, new productivity and usability challenges have also emerged. A commonly overlooked limitation of current systems is that they do not provide effective integration of analytics and end user workspace environments (e.g., investigators, analysts user productivity tools).  We discuss critical challenges in the effective integration of data-driven insights and end user- oriented case management technologies. We discuss synergies between micro-services, analytics, case management as step forward to empower end users to effectively use big data technologies, while share and collaborate on the fly, in order to generate and evolve insights.

Bio: Prof. Boualem Benatallah is a Scientia Professor and research group leader at the School of Computer Science (CSE), University of New South Wales (UNSW, Sydney, Australia). He held the chaire d’excellence position of the Auvergne Region in France (LIMOS, France, 2008-2010). His main research interests are developing fundamental concepts and techniques in service composition, cloud service engineering, end user analytics, quality control in crowd sourcing systems, and business processes management. He has published more than 200 refereed papers including more than 60 journal papers. Most of his papers appeared in very selective and reputable conferences and journals. He has been frequently invited to give keynote talks, lectures and tutorials in international conferences and summer schools. He has a very strong international track record demonstrated by the high citations of his work, some of which are considered seminal in the field of services composition. Boualem has been PC co-chair of number of international conferences including BPM’05, ICSOC’05, WISE’07, ICWE’2010, IEEE/ACM WI’11, IEEE SOCA’11. He was research track co-chair for the WWW’11 conference. He was the general chair of ICSOC’08 – Sydney. He has acted as a key official (tutorial chair, workshops chair, publication chair, area chair, PhD symposium chair) for several international conferences. He has been guest editor of number of special issues for reputable international journals including ACM TOIT. He has been a PC member of all the reputable international conferences in his areas of research including VLDB, ICDE, WWW, EDBT, MDM, ICSOC, ICWS and ER. He is member of the steering committee of BPM and ICSOC conferences. He is member of the editorial board of numerous international journals and series including ACM Transactions on Web and IEEE transactions on Cloud Computing. He was a visiting Professor at INRIA-LORIA, CNRS, Claude Bernard University (France), University of Blaise Pascal (Clermont Ferrand, France), University of Trento (Italy), University of Paris-Dhauphine (France). As the chair of the CSE research committee, he was member of the team (comprising multiple university, government and industry partners) that constructed the successful bid for the Smart Services CRC (Cooperative Research Centre) in 2007. He is a leader of the data curation research stream at the new Data to Decisions CRC. He is member of Executive Committee of IEEE Computer Society’s Technical Committee on Business Informatics and Systems and Australian Service Science Society.

Value Creation from Big Data Analytics using IBM Watson

Presenter(s): Karthikeyan Ponnalagu – IBM

Date: September 20, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Why NP-completeness matters and how

Presenter(s): Prof. Aditya Ghose – UOW

Date: August 18, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: This is a popular science talk, where I will try to revive what appears to be flagging interest in theoretical computer science. My main focus will be on explaining Cook’s Theorem and its proof, but along the way, I will also address interesting questions arising from human computation, social computing and the connections between literature and the development of computational artefacts.

Real-valued distributed constraint optimization

Presenter(s): Graham Billiau – UOW

Date: August 11, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Big Data Framework for Agile Business (BDFAB v2.0)

Presenter(s): Dr. Bhuvan Unhelkar – Associate Professor (IT) at University of South Florida Sarasota-Manatee (USFSM) in Florida and Founding Consultant at MethodScience

Date: July 28, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: The term “big data” encompasses a wide variety of topics led by the two keywords “analytics” and “technologies. Technically, big data implies Hadoop/HDFS,  Spark and, at the back end, NoSQL. From a business viewpoint, however, big data analytics command  greater interest  as they enable identification of patterns, facilitate predictions, and also provide prescriptive advise for better decision making. When medium to large enterprises want to adopt big data, they need to go through the rigors of large-scale adoption through people, processes, and technologies. Their Big data analytics need to be coupled with a proper understanding of technological capabilities provided by  enterprise architectures. We thus find that there are multiple technical, analytical and architectural elements that come into play in big data adoption by organizations. This discussion focuses on an overarching framework that will not only facilitate adoption of analytics and technologies but will provide a solid foundation for a strategic approach to big data. This framework is called the Big Data Framework for Agile Business (BDFAB v 2.0). The key elements composing this framework include: Agile values for business, organizational roles in big data, building blocks of big data strategies for business (including the role of analytics within those strategies), key artifacts in big data adoption, business conditions and limitations, Agile practices, and a compendium (repository). The building blocks of big data strategies are themselves are made up of five modules:

  1. Business decisions,
  2. Data & technology
  3. Optimization & Visualization
  4. Organizational Capabilities and EA
  5. People (capabilities) & Quality

In addition, this framework is accompanied by a 12-lane process for big data transformation especially in large organizations. Exploration of this framework will be of practical benefit to organizations looking for a sensible pathway into big data and, at the same time, provide opportunity for refinement based on further experimentation.
BDFAB is the core contribution of a book I am publishing with CRC press (USA) on Big Data Strategies (due 2016 end).

Bio: Dr Bhuvan Unhelkar (BE, MDBA, MSc, PhD; FACS; CBAP) has more than two decades of strategic as well as hands-on professional experience in the field of Information and Communication Technologies (ICT) industry. As Associate Professor (IT) he leads the IT program at University of South Florida, Sarasota-Manatee campus; and as Founding Consultant at MethodScience he has developed mastery in Business Analysis & Requirements Modeling, Software Engineering, Big Data Strategies, Agile Processes, Mobile Business and Green IT. His domain experience is banking, financial, insurance, government and telecommunications. Bhuvan is a thought-leader, an author (17 books – latest being The Art of Practicing Agile: Applying Composite Agile Method and Strategy to Real Projects (Taylor and Francis USA). and a coach and a trainer. Recent Cutter executive reports (Boston, USA) include Psychology of Agile, Business Transformation, Collaborative Business & Enterprise Agility and Agile in Practice-a Composite approach. He is a winner of the Computerworld Object Developer Award (1995), Consensus IT Professional Award (2006) and IT Writer Award (2010). He has a Doctorate in the area of “Object Orientation” from the University of Technology, Sydney, in 1997. He was Sr. Lecturer at Western Sydney University and an adjunct/guest professor at the University of Technology, Sydney; and the Australian Catholic University. Bhuvan is Fellow of the Australian Computer Society, Life member of Computer Society of India and Baroda Management Association, Member of SDPS, Rotarian at St.Ives (Past President & Paul Harris Fellow), Discovery volunteer at NSW parks and wildlife, and a previous TiE Mentor.

Predicting Database Workloads through Mining Periodic Patterns in Database Audit Trails

Presenter(s): Dr Janusz Getta – UOW

Date: July 21, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: Information about the periodic changes of intensity and structure of database workloads plays an important role in performance tuning of functional components of database systems. Discovering the patterns in workload information such as audit trails, traces of user applications, sequences of dynamic performance views, process logs, etc. is a complex and time consuming task. In the presentation we shall investigate a new approach to analysis of information included in the database audit trails. In particular, we shall present how to transform information included in the audit trails into a format that can be used for discovering the periodic patterns in the fluctuations of database workloads. We shall present few models of periodic patterns and the algorithms for discovering periodic patterns in audit trails. We shall also present a collection of rules that can be used to derive  complex periodic patterns from the elementary and other complex patterns and and we shall show how to use such rules to predict the future workload levels.

Deep learning in non-cognitive domains

Presenter(s): Dr Truyen Tran – Deakin University

Date: July 13, 2016
Time: 10am onwards
Venue: 6.312 – Smart Building

Abstract: The current wave of deep learning has broken many records long held by traditional machine learning. The main power of deep learning is to deliver end-to-end systems that learn from raw data, extract multiple feature abstractions, and predict outcomes with little human intervention. All of the successes thus far are in cognitive domains such as vision, speech and languages, where humans do well with little effort. In this talk, I present our recent deep learning work in other domains that require specialized training for humans to perform reasonably, including healthcare, software engineering, anomaly detection, multi-relational databases, and mixed-type analysis.

Bio: Truyen Tran is a lecturer at Deakin University. He works in broad range of data analytics fields with current focus in deep learning, healthcare and software engineering. He received multiple paper awards and prizes including UAI 2009, CRESP 2014, Kaggle 2014 (where he is Kaggle Master), PAKDD 2015, and ACM SIGSOFT 2015. He obtained a Bachelor of Science from University of Melbourne and a PhD in Computer Science from Curtin University in 2001 and 2008, respectively.

Cloud-driven Participatory Sensing for Urban Informatics

Presenter(s): Tridib Mukherjee – Senior Research Scientist, Xerox Research Center India (XRCI)

Date: June 30, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: The rapid advancements in sensing, computation and communications have led to the proliferation of smart phones. People-centric sensing is a scientific paradigm which empowers citizens with sensor-embedded smartphones, to contribute to micro and macro-scale urban sensing applications – either implicitly (in an opportunistic manner) or explicitly (in a participatory manner). Community-based urban sensing applications, are typically participatory in nature. For instance, commuters reporting on a transit overload may explicitly need to provide an input through an app to report on the overload.  This talk will focus on the trends, challenges and applications of participatory sensing systems. Additionally, the talk will cover the solution requirements for effective deployments of such systems in real scenarios. The solution approach cuts across a multiple computer science disciplines such as, large-scale data management, cloud computing, platform design principles, and data analytics & mining.

Bio: Tridib Mukherjee is a Senior Research Scientist at the Xerox Research Center India (XRCI). He works in the broad areas of Distributed Computing, Cloud Computing, Large-scale Systems, Green Computing, Sensor Networks, Mobile Computing, and Services Computing. He joined XRCI in September 2011 and has since been involved in exploring research and business opportunities in efficient delivery of SaaS offerings and other enterprise applications. Tridib’s work on CloudAdvisor, a cloud configuration recommendation system for enterprise application, won the prestigious IEEE Cloud Cup in 2013. He is currently working on platform optimizations in enterprise clouds for efficient service management as well as for large scale mobile, participatory, & pervasive sensing, data integration & analytics (especially for smarter city applications). Prior to joining Xerox, Tridib was a Postdoctoral Research Scholar at the Arizona State University where he worked on data center optimization and body sensor networks. Tridib has co-authored a book, multiple book chapters, and published over 50 papers in reputed journals and conferences. Tridib also has 3 granted US patents in addition to more than 30 US patents filed.

Predicting Delivery Capability in Iterative and Agile Software Development

Presenter(s): Morakot Choetkiertikul – UOW

Date: June 16, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: Agile and iterative software development has become widely practiced in the industry. Since agile projects require fast, incremental delivery in every iteration of software development, it is essential to monitor the execution of an iteration, and predict and manage relevant risks as the iteration progresses. We present a novel, data-driven approach to providing automated support for project managers and other decision makers in predicting delivery capability for an ongoing iteration. Our approach leverages a history of project iterations and associated issues, and in particular, we extract characteristics of previous iterations and their issues in the form of features. In addition, our approach characterizes an iteration using a novel combination of techniques including feature aggregation statistics, automatic feature learning using the Bag-of-Words approach, and graph-based complexity measures.  An extensive evaluation on seven large open source projects demonstrates the accuracy of our predictive models, achieving on average 75% precision, 72% recall, 72% F-measure, 89% Area Under the ROC Curve and a low error rate of 0.36 for Macro-averaged Mean Absolute Error.

Optimisation of Query Processing with Multilevel Storage
(Joint work with Dr. Janusz R. Getta, DSL)
(Paper already presented at the CIIDS-2016 conference)

Presenter(s): Nan Noon Noon – UOW

Date: May 19, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: The typical algorithms for optimization of query processing in database systems do not take under the consideration the availability of different types and sizes of persistent and transient storage resource that can be used to speed up the internal query processing. It is well known that appropriate allocation of storage resources for the internal query processing may significantly improve performance. This paper describes the new algorithms for automatic management of multilevel transient and persistent storage resources in order to optimize the performance of query processing in a database system.  The algorithms presented in the paper process the concurrently submitted queries and discover the common query processing plans. The algorithms estimate the query processing costs and choose the best allocation of multilevel storage resources to optimise the overall internal query processing costs.  The paper presents the outcomes of experiments that confirm the improvements in performance through appropriate allocation of multilevel storage for the internal query processing.

Scalable Stochastic Models for Cloud Services

Presenter(s): Rahul Ghosh, PhD – Research Scientist, Xerox Research Center India

Date: May 13, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: In this talk, we present high fidelity stochastic models for enterprise systems using Cloud computing as an example. Large scale Cloud services are inherently stochastic in nature. Given the large number of system states, any naive modeling approach quickly becomes intractable and scale of the Cloud makes the problems challenging in terms developing practical solutions. Hence, quantifying and predicting the service quality (e.g., performance and dependability) of such Clouds is of interest. This presentation will deep dive into scalable performance, power-consumption and resource over-commit analysis for an Infrastructure-as-a-Service (IaaS) Cloud. The talk will also briefly cover other what-if analysis approaches such as Cloud downtime analysis, capacity planning and resiliency quantification. Models and approaches presented in this research can be useful to a Cloud service provider for planning, forecasting, bottleneck detection, what-if analysis or overall optimization during design, development, testing and operational phases of a Cloud.

Bio: Rahul Ghosh is a Research Scientist at Xerox Research Center India. He received his M.S. and Ph.D. in Electrical and Computer Engineering from Duke University, USA in 2009 and 2012 respectively. Prior to this, he received his B.E. in Electronics and Telecommunication Engineering from Jadavpur University, India in 2007. Before joining Xerox, Rahul was an Advisory Software Engineer at IBM Research Triangle Park, USA. Besides being the reliability architect for an IBM Cloud product, he also drove performance and quality improvements in Cloud. Rahul’s research interests include services research, performance and dependability analysis of large scale computer systems, and stochastic processes. During his Ph.D., he also worked as a research intern at IBM T.J. Watson Research Center. Rahul is a co-author of 25+ peer-reviewed conference and journal papers and has filed 20+ US patents (2 issued) primarily in Cloud and analytics.

Towards Cloud-based Decentralized Storage for Internet of Things Data

Presenter(s): Dr. Nanjangud C Narendra – Principal Engineer, Ericsson Research Bangalore

Date: March 31, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: The Internet of Things (IoT) phenomenon is creating a world of billions of connected devices generating enormous amounts of data. That data needs to be stored efficiently so that it can be retrieved easily on demand and acted upon. Current cloud-based solutions, focusing on centralized data collection and storage, would not be adequate for this task, given the sheer volume of data expected to be generated by IoT devices. Despite this, however, very little work has been reported on cloud-based storage tailored for IoT data. To that end, in this paper, we present a decentralized cloud-based storage solution specifically tailored for IoT data. The salient features of our solution are: usage of object storage (such as Ceph) for software-defined storage; and optimal distribution of data among distributed mini-Clouds, which are mini-data centers. We present approaches for the following: optimal mini-Cloud placement to minimize latency of data collection from IoT devices; and data migration among mini-Clouds with a view towards addressing storage capacity issues while minimizing access latency. Throughout our paper, we illustrate our ideas via a realistic running example in the Smart Cities domain, and present experimental results via a proof of concept prototype.

Bio: Dr. Nanjangud C Narendra (NCN) has joined Ericsson Research Bangalore as a Principal Engineer. Prior to joining Ericsson, he worked in MS Ramaiah University, Cognizant, IBM Research, HP India and Motorola India. He has about 22 years R&D experience in the Indian IT industry. His research interests span Software Engineering, Workflow Management, Web Services, Service Oriented Architecture, Cloud Computing and Internet of Things. He has published over 100 papers in international conferences and journals. He has been Program Committee member for several key conferences such as Autonomous Agents and Multi-Agent Systems (AAMAS), International Conference on Service Oriented Computing (ICSOC), IEEE International Cloud Computing Conference; he was also Program Committee for ICSOC (Industry track) in 2012 and ICSOC (Research track) in 2015. He is a member of the Editorial Board of Service Oriented Computing and Applications journal. He is a Senior Member of IEEE and ACM.

Goals, games and business processes (CANCELLED)

Presenter(s): Prof. Aditya Ghose – UOW

Date: March 24, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Social Specifications of Business Processes with Azzurra

Presenter(s): Prof. John Mylopoulos – University of Toronto (Canada) and University of Trento (Italy)

Date: February 8, 2016
Time: 1.30pm onwards
Venue: 6.209 – Smart Building

Abstract: A business process is first and foremost a social interaction among multiple participants. Business process modeling languages support the description of business processes in operational terms, as collections of interleaved activities conducted by human and software agents. However, such descriptions do not capture adequately the richness of social interaction among participants. To address this deficiency, we propose Azzurra, a specification language for modeling and enacting business processes. Azzurra is founded on social concepts, such as roles, agents and commitments among them, and Azzurra specifications are social models consisting of sets of commitments. As such, Azzurra specifications support flexible executions of business processes, and provide a semantic notion of actor accountability and business process compliance. In this talk, we present syntax and semantics of Azzurra, and we sketch algorithms to determine runtime compliance with an Azzurra specification.

Bio: John P. Mylopoulos is professor at the University of Toronto, Canada, and at the University of Trento, Italy. He is known for his work in the field of conceptual modeling, specifically the development an agent-oriented software development methodology called TROPOS.
Mylopoulos was awarded the Peter P. Chen Award for outstanding contributions to the field of conceptual modeling in 2010. In 2012 he also received a Honorary Doctorate from the RWTH Aachen University in recognition of “his excellent and distinctive contributions on the methodology of conceptual modeling as a basis for databases, software technology and artificial intelligence, as well as its interdisciplinary applications.” He is a fellow of the American Association for AI (AAAI) and the elected president of the VLDB Endowment (1998-01, re-elected for the period 2002-05). He has served as editor of the Requirements Engineering Journal, published by Springer-Verlag.

Semantic Process Annotation in ProcessSEER

Presenter(s): Kerry Hinge – UOW

Date: March 3, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Controlled violation of temporal process constraints – models, algorithms and results

Presenter(s): Professor Akhil Kumar – Smeal College of Business, Penn State University, University Park, PA 16802, USA

Date: January 12, 2016
Time: 4pm onwards
Venue: 6.105 – Smart Building

Abstract: While there has been much work on modeling and analysis of temporal constraints in workflows in the context of many real-world applications, there has not been much work on managing violations of temporal constraints. In real-time workflows, such as in medical processes and emergency situations, and also in logistics, finance and in other business processes with deadlines some violations are unavoidable.  Here we introduce the notion of controlled violations as the ability to monitor a running process and develop an approach based on constraint satisfaction to determine the best schedule for its completion in a way so as to minimize the total penalty from the violations.  The violations are evaluated in terms of metrics like number of violations, delay in process completion, and penalty of weighted violations. We also relate our work to the concept of controllability in literature and show how it can be checked using our method. Finally, the expressive power of our approach is discussed and we also propose a spreadsheet based heuristic approach for managing temporal constraints.

Bio: Akhil Kumar is a professor of Information Systems at Penn State. He received his Ph.D. from Berkeley and has previously been on the faculties at Cornell University and University of Colorado. He also spent a sabbatical year as a scientist at Bell Labs, Murray Hill, NJ. He has published more than 100 papers in academic journals and international conferences. He was coauthor of a paper which received the best paper award at the INFORMS Conference on Information Systems and Technology in November 2001.
Akhil currently serves as an associate editor for ACM Transactions on Management Information Systems. He is also on the editorial board of Information and Technology Management Journal. He has previously been an associate editor for IEEE Transactions on Services Computing, INFORMS Journal on Computingand Information Systems Research. He also served as a co-program chair of CoopIS’11 and a co-chair of Workshop on Information Technologies and Systems (WITS’07). He has been a principal investigator for National Science Foundation, and also received support from IBM, Sun Microsystems and other organizations for his work.

His current research interests are in BPM and workflow systems, health IT, service composition, process mining and business analytics.