Description du projet
Raisonnement solide pour les moteurs de recherche sémantique
Une simple recherche d’informations sur l’internet n’est pas toujours simple. Il s’agit souvent de passer au crible des résultats de recherche non pertinents. L’un des problèmes des moteurs de recherche sémantique actuels est qu’ils reposent sur l’hypothèse irréaliste que tous les faits réels concernant un domaine donné sont explicitement énoncés dans leur base de connaissances ou sur le web. Les informations pertinentes sont souvent réparties dans plusieurs bases de connaissances incohérentes, et les théories du domaine sont rarement complètes. Financé par le Conseil européen de la recherche, le projet FLEXILOG vise à améliorer les moteurs de recherche sémantique en introduisant une famille de logiques pour un raisonnement solide avec des connaissances du monde réel. Il utilisera des représentations de l’espace vectoriel des termes du langage naturel pour estimer la plausibilité des modèles logiques, permettant ainsi diverses formes de raisonnement de bon sens.
Objectif
Semantic search engines use structured knowledge to improve traditional web search, e.g. by directly answering questions from users. Current approaches to semantic search rely on the unrealistic assumption that all true facts about a given domain are explicitly stated in their knowledge base or on the web. To reach their full potential, semantic search engines need the ability to reason about known facts. However, existing logics cannot adequately deal with the imperfect nature of knowledge from the web. One problem is that relevant information tends to be distributed over several heterogeneous knowledge bases that are inconsistent with each other. Moreover, domain theories are seldom complete, which means that a form of so-called plausible reasoning is needed. Finally, as relevant logical theories do not exist for many domains, reasoning may need to rely on imperfect probabilistic theories that have been learned from the web.
To overcome these challenges, FLEXILOG will introduce a family of logics for robust reasoning with messy real-world knowledge, based on vector-space representations of natural language terms (i.e. of lexical knowledge). In particular, we will use lexical knowledge to estimate the plausibility of logical models, using conceptual simplicity as a proxy for plausibility (i.e. Occam’s razor). This will enable us to implement various forms of commonsense reasoning, equipping classical logic with the ability to draw plausible conclusions based on regularities that are observed in a knowledge base. We will then generalise our approach to probabilistic logics, and show how we can use the resulting lexically informed probabilistic logics to learn accurate and comprehensive domain theories from the web. This project will enable a robust data-driven approach to logic-based semantic search, and more generally lead to fundamental progress in a variety of knowledge-intensive applications for which logical inference has traditionally been too brittle.
Champ scientifique
- natural sciencescomputer and information sciencesknowledge engineeringontology
- natural sciencesmathematicsapplied mathematicsstatistics and probabilitybayesian statistics
- natural sciencescomputer and information sciencesdata sciencenatural language processing
- natural sciencescomputer and information sciencesartificial intelligencemachine learning
- natural sciencescomputer and information sciencesartificial intelligencecomputational intelligence
Programme(s)
Thème(s)
Régime de financement
ERC-STG - Starting GrantInstitution d’accueil
CF24 0DE Cardiff
Royaume-Uni