Structural and semantic proximity in information networks

This research includes the investigation, design and experimentation of models and measures of semantic and structural proximity for knowledge extraction and link prediction. The aim is to measure, predict and elicit, in particular, data from social or collaborative sources of heterogeneous information. The general idea is to use the information about entities (i.e. users) and relationships in collaborative or social repositories as an information source to infer the semantic context, and relations among the heterogeneous multimedia objects of any kind to extract the relevant structural knowledge. Contexts can then be used to narrow the domains and improve the performances of tasks such as disambiguation of entities, query expansion, emotion recognition and multimedia retrieval, just to mention a few. There is thus the need for techniques able to produce results, even approximated, with respect to a given query, for ranking a set of promising candidates. Tools to reach the rich information already exist: web search engines, which results can be calculated with web-based proximity measures. Semantic proximity is used to compute attributes e.g. textual information. On the other hand, non-textual (i.e. structural, topological) information in collaborative or social repositories is used in contexts where the object is located. Both web-based and structural similarity measures can make profit from suboptimal results of computations. Which measure to use, and how to optimize the extraction and the utility of the extracted information, are the open issues that we address in our work.