Short description of her research activity
The research activity of Gabriella Pasi is finalized at defining innovative models and systems that allow an effective access to digital information relevant to specific user needs. Since a few years, she is also working on the issue of analyzing various aspects of the information (user generated content) produced on social media.
In relation to the issue of accessing information relevant to specific users’ needs, the considered classes of systems include Information Retrieval (Search Engines), Information Filtering, Database Management Systems and Decision Support Systems. The common objective of the research activities undertaken in relation to this issue is defining flexible and intelligent systems, able to manage both precise and imperfect information, i.e., information affected by imprecision/vagueness and/or uncertainty, and to adapt to the user context.
The design of such flexible systems has been pursued by investigating the application of “soft computing” techniques (mainly based on fuzzy set theory, possibility theory, and fuzzy multi-criteria decision making) to various tasks concerning both information representation and retrieval. In particular, the use of Possibility Theory has been investigated to define models for the representation of incomplete data and information; Fuzzy Set Theory has been applied to define both flexible query languages, and innovative retrieval models.
Multi Criteria Decision Theory, especially concerning the problem of aggregation, has been applied in both contexts of IR (e.g., for multidimensional relevance assessment) and of social media analytics.
She is actually working on various research topics related to the above issues, among which:
- Definition of systems for contextual information access
- Context modelling
- User Profiling by means of: language and topic models, semantic networks, deep neural networks
- Social media analytics
- Analysis of information diffusion on social networks