Séminaire AMA dans le cadre du recrutement LIG sur le poste MCF/Chaire CNRS "Web et réseaux sociaux"

- Romain Boulet

• Théorie algébrique de graphes, analyses de réseaux réels interdisciplinaires

  J’exposerai dans un premier temps mes travaux en théorie spectrale de graphes dont un des buts est d’établir des liens entre structure d’un graphe et ses éléments spectraux (valeurs propres et/ou vecteurs propres de matrices associées au graphe). Je présenterai ensuite l’étude d’un réseau social entre paysans au Moyen Age mêlant outils algébriques et statistiques puis un réseau de liens entre textes juridiques. J’aborderai également le domaine de la recherche d’information via l’analyse de thésaurus modélisés sous forme de graphes. Je terminerai par l’exposé de perspectives liées à ces travaux dans le domaine de l’analyse de réseaux sociaux et de la recherche d’information.

- Amin Mantrach

• A General Framework for People Retrieval in Social Media with Multiple Roles

  Abstract : Internet users are more and more playing multiple roles when connected on the Web, such as "posting", "commenting", "tagging" and "sharing" different kinds of information on various social media. Despite the research interest in the field of social networks, few has been done up to now w.r.t. information access in multi-relational social networks where queries can be multifaceted queries (e.g. a mix of textual key-words and key-persons in some social context). We propose a unified and efficient framework to address such complex queries on multi-modal ``social’’ collections, working in 3 distinct phases, namely : (I) aggregation of documents into modal profiles, (II) expansion of mono-modal subqueries to mono-modal and multi-modal subqueries, (III) relevance score computation through late fusion of the different similarities deduced from profiles and subqueries obtained during the first two phases. Experiments on the ENRON email collection for a recipient proposal task show that competitive results can be obtained using the proposed framework.

- Anna Stavrianou

• Extracting interesting information from social media comments

  Discussions and comments in social media encapsulate people’s interests and beliefs. Mining and extracting quality knowledge from such discussions is significant, yet, not straightforward. In this talk the focus is on web forums as well as online reviews. We discuss about a graph-oriented model, accompanied by a number of structure and content-oriented measures which aim at facilitating browsing, analysis and extraction of interesting postings from forums. The talk will continue with a discussion of how personalization techniques could enhance the extraction task and how the extracted information could be used within recommender systems.