My main research activities spin around the following themes.
Knowledge-based explainability and transparency of AI and ML models
Transparency and explainability in image classification are essential for establishing trust in machine learning models and detecting biases and errors. State-of-the-art explainability methods generate saliency maps for images or other types of content to show where a specific class or result is identified. My research focuses on post-hoc methods that combine knowlege graphs, concept-based approaches and human-AI interaction methods to maximize the understandability and acceptability of AI / ML model results for users.
Interaction Flow Modeling Language (IFML) for UI Modeling
I’m one of the co-inventors of IFML (Interaction Flow Modeling Language), an international standard adopted by the OMG in 2014, and of its ancestor WebML (Web Modeling Language). This work lead to several scientific publications and also to a monographs published by Morgan Kauffmann in 2002 on WebML (the cornerstone of the Web Engineering discipline, with currently 617 citations) and in 2015 on IFML. The contribution to the field is also demonstrated by the widespread adoption of the associated tool WebRatio, now used by more than 17,000 students and researchers throughout the world for learning and research purposes. I have been the main promoter and coordinator of the IFML standardization.
Model-driven Engineering
My research activities within the field of MDE (Model Driven Engineering) consisted in the study of new software modeling approaches. My most recently publication is the book Model Driven Software Engineering in Practice (Morgan & Claypool, USA, 2012, co-authored with two well-known European researchers in the MDE field: Jordi Cabot from EMN, France, and Manuel Wimmer from TUW, Austria; the book is currently in its second edition, published by Springer) that is currently the reference book for MDD / MDE (currently adopted by more than 100 universities worldwide). I have published several papers and demonstrated and validated various prototype tools in this field.
Exploratory Multidomain Search and Crowdsourcing
My research focuses also on information management and advanced Web search, spanning multi-domain and exploratory search too (work done within Pharos, Cubrik and SeCo projects). These activities already produced some results and relevant publications on: exploratory search paradigms (WWW Conf. 2010, VLDB Journal), crowdsourcing systems for information seeking (WWW 2012) and their evolution towards reactive systems (WWW 2013). I am co-author of two patents on these topics and of a book on Web Information Retrieval published by Springer in 2014.
Data Science and Social Media Analysis
More recently, I have broadened my research on data analysis to big data problems involved in the analysis of large scale social media and real world events. I study data fusion, wrangling and analysis on cross-disciplinary projects related to smart cities, large scale events, and social network user behavior and trends.