Panel on Future Trends of Software Engineering and Knowledge Engineering

Du Zhang
California State University, USA
(Moderator)

The International Conference on Software Engineering and Knowledge Engineering (SEKE) is celebrating its 24th anniversary this year. For nearly a quarter of century, while SEKE has established itself as a major international forum to promote research and practice in software engineering and knowledge engineering, the computing fields have undergone profound changes. Today, our daily lives are intimately intertwined with artifacts that are the results of software engineering and knowledge engineering. What will the future hold for SEKE as a field of inquiry in the next ten years? What are the challenges that lie ahead? What can we do as a community to further our agenda on SEKE? Toward illuminating our path to the future, an excellent panel of experts has been assembled. Panelists will share their insight on the future trends of software engineering and knowledge engineering. We hope you will find the panel an inspiring impetus for the continued growth of SEKE in the years to come.

Software Engineering of Autonomic Clouds

Masoud Sadjadi
Florida International University, USA
(Panelist)

Autonomic or self-managing clouds are becoming prevalent software deployment environments for applications ranging from commerce (e.g., banking), to education (e.g., virtual labs), to research (e.g., high-performance computing). Unfortunately, traditional approaches to software engineering are not applicable to the specific characteristics of autonomic clouds, which are becoming a major part of every software application's solution domain. Therefore, there is a desperate need for a paradigm shift in how software applications are designed, developed, tested, deployed, hosted, and consumed in the clouds. One example of the specific characteristics of autonomic clouds is the concept of on-demand services leasing, which has major impacts on the growth of new businesses, from their inception to booming popularity. To respond to such needs, service providers face major challenges when trying to keep up with their promise of infinite capacity with unconditional elasticity.

Big Data in Software Engineering: Challenges and Opportunities

Taghi Khoshgoftaar
Florida Atlantic University, USA
(Panelist)

The field of software engineering has changed drastically in the past 20 years. Although traditional quality assurance approaches such as unit tests and change tracking remain essential tools, these approaches can be easily overwhelmed by the sheer volume of modules, bugs, programmers, and projects managed in large software development firms. To deal with this "Big Data," a new class of software engineering tools are needed: those from the fields of data mining and machine learning. By employing techniques specifically designed to sift through enormous datasets and identify the elements in need of human attention, data mining tools permit software practitioners to focus valuable human resources where they are needed most. I will discuss a number of topics concerning the use of data mining to manage Big Data in the context of software engineering, including software metric selection, data balance issues, and quality of data.

Knowledge Engineering, Operational Research and AI: the Time to Meet

Eric Gregoire
Universite d'Artois, France
(Panelist)

Although they share many paradigms, the Operational Research and Artificial Intelligence fields have often evolved separately. This last decade, both domains have come ever closer, through new insights in constraint solving and SAT-related technologies, allowing problems to be solved that were long considered out of reach. This opens new perspectives for Knowledge Engineering as well.

Computational Issues in Social Networks

Swapna S. Gokhale
University of Connecticut, USA
(Panelist)

Online social networks (OSNs) have had an enormous impact on the way people communicate and share information. Today, the population of Facebook exceeds that of the United States and Lady Gaga has more Twitter followers than the entire population of Australia! OSNs not only provide social channels for communication, but they also offer critical marketing and customer profiling tools for businesses. This revolution has precipitated a deep desire to understand the structure of OSNs, identify the latent patterns that may exist within these networks, and leverage these structures and patterns to build novel applications and services. While sociologists have researched such social networks for decades, never before has such a vast quantity of structured social network data xxviii been available for analysis. Social network analysis is thus a rapidly emerging field that combines algorithmic, graph theoretical, and data mining techniques to map, measure, and find patterns in the relationships and communication flows in massive OSN datasets. This talk will summarize the recent, state-of-the-art research in OSN analysis on topics such as topology characterization, information and influence diffusion, community detection, inferring relationship strength, microblog analysis, friend and link prediction, data anonymity, workload characterization, and security and privacy, and outline avenues for further exploration.

Web Intelligence: Representation and Processing of Knowledge with Uncertainty

Marek Reformat
University of Calgary, Canada
(Panelist)

Uncertainty is an integral component of information and knowledge. Many concepts we deal with are without precise definitions, or with unknown facts, missing or inaccurate data. Such a situation is also present on the Internet where many sources of information could be corrupted, or partially and temporally inaccessible. Our dependence on the Internet is growing with every day. We relay on it doing research, learning new things, and finding what is happening in the world and in our neighborhood. But, how much imprecision and ambiguity is out there? How many sources of data are trustworthy? How much we can relay on the web to discover new things? Additionally, uncertainty is not only associated with data and information stored on the web . the users also bring ambiguity and imprecision. In many cases, the users' behavior and decisions depend on current circumstances, users' judgments, their understanding of situations, and their needs and requirements . things that are .equipped. with ambiguity.

In order to make the web a user-friendly environment where the users can easily and quickly find things they are looking for, new web utilization tools have to be developed. They should be able to deal with numerous alternatives provided by the Internet, as well as with imprecision.

The purpose of this topic is to provoke discussion how critical is to address the issue of imprecision and what methods, tools and approaches would be possible solutions.