Ponencia JISBD 2019 – 2 de septiembre de 2019
Ponente: Dr. Paris Avgeriou
Dr. Paris Avgeriou is Professor of Software Engineering at the University of Groningen, the Netherlands where he has led the Software Engineering research group since September 2006. Before joining Groningen, he was a post-doctoral Fellow of the European Research Consortium for Informatics and Mathematics. He is the Editor in Chief of the Journal of Systems and Software, as well as an Associate Editor for IEEE Software. He also sits on the board of the Dutch National Association for Software Engineering (VERSEN) and the Dutch research school IPA. He has co-organized several international conferences such as ECSA, ICSA, and Tech Debt and served on their steering committees. His research interests lie in the area of software architecture, with strong emphasis on architecture modeling, knowledge, evolution, analytics and technical debt. He champions the evidence-based paradigm in Software Engineering research and works towards closing the gap between industry and academia.
Technical Debt: talking about the elephant in the room
The term Technical Debt has undeniably become part of the everyday vocabulary of software engineers. We know that it concerns compromises to the internal quality of a system, made either deliberately or inadvertently. We understand that it’s not all bad, as it may have served the purpose of expediency at some point. But, it is associated with a clear risk especially for large and complex systems: if we do not manage Technical Debt, it threatens to “bankrupt” those systems. Action must be taken before we reach the point of no return. In this talk we revisit the state of the art in managing Technical Debt, we identify challenges and discuss promising future directions. We pay special attention to Architecture Technical Debt, which requires the most effort during maintenance and evolution but is often overlooked by tool vendors.
Ponencia PROLE 2019 – 3 de septiembre de 2019
Ponente: Fausto Spoto (JuliaSoft, Verona, Italia)
Fausto Spoto holds a PhD in Computer Science from the University of Pisa (Italy) and is an Associate professor at the University of Verona (Italy). He started his research career by investigating on static analysis for declarative programming languages. Then, he developed similar techniques for the static analysis of object-oriented programs, by abstract interpretation. In 2003, he started developing the static analyzer Julia for Java bytecode and later founded JuliaSoft Srl, a company whose goal is the commercialization of Julia. He’s been the main researcher and developer of the tool in JuliaSoft Srl, until 2018, when he left the company.
Ponencia JCIS 2019 – 4 de septiembre de 2019
Ponente: Fabio Casati
Fabio Casati is a professor of software engineering and machine learning at the University of Trento. Until 2006, he was technical lead for the research program on business process intelligence in Hewlett-Packard USA, where he contributed to several HP commercial products in the area of web services and business process management. He then moved to academia, where he started research lines on hybrid human-machine computations and on technologies for happiness and life participation, focusing on achieving direct positive impact on society through tangible artefacts, widely adopted by the community. He is co-author of a best-selling book on Web services and author of over 250 peer-reviewed papers.
Can Crowd and Machines Help us Understand and Summarize Scientific Knowledge? The case of Literature Reviews
Literature reviews and meta-analyses are one of the most useful scientific activities – and one of the main form of publications in science. They are often highly cited, and form the basis for evidence-based practices and even government policies, from education to healthcare, as they pool insights and results independently obtained from a number of research groups. They are also very helpful in introducing newcomers to challenges and opportunities for research in a given area.
Reviews, especially when done scientifically and systematically, may however also be complex, lengthy, and frustrating. They are also often marred by quality issues and, with millions of papers published every year, they face the risk of being out of date as soon as we are done writing them.
In our work we explore if and how the “crowd” and machines, through AI, can assist scientists in rapidly obtaining high quality literature review at a very low cost, while still retaining not only the informative value and scientific validity of the reviews but also the educational value (e.g., for PhD students) of going through the review process.
I will show that task design and crowdsourcing algorithms are central to the solution even when no AI is involved. I will then present how crowd and AI can be combined – in different ways for each different problem, because each review is unique in many ways – to achieve even better results.
Finally, I’ll briefly discuss how this line of work can be extended to support all fields where evidence-based approaches are (or should be) sought.