Hajer Baazaoui-Zghal

Modular ontologies and CBR-based hybrid system for web information retrieval

Increasing amounts of data volume used on the Web and their heterogeneous character make the search for information a challenging task. Several advanced computing methods and technologies propose to incorporate a degree of semantic analysis during the search based on ontologies. Ontology engineering is based on the methods and methodologies for building ontologies combined with engineering process. This paper proposes a hybrid system based on ontology engineering and aiming to enhance web information retrieval results, by combining automatic Modular Ontology building with CBR (CBRModOnto). The system integrates a novel dynamic composition approach of modular ontologies, performed to reorganize the overlapping concepts between the different ontology modules and updates their hierarchy using semantic similarity measure. The search in our system occurs in two main phases. The first one composes modular ontologies and manages the knowledge base, while the second phase manages the CBR process. A demonstration case study presents a scenario to illustrate the proposed system. Our system has been implemented, the obtained experimental results show that hybridization that we propose enables an improvement of query reformulation, predicted ranking score and user’s satistaction.

An Ontology-Driven Visual Question-Answering Framework

Question Answering systems aim at providing answers to natural language questions and provide a solution to the problem of response accuracy. This paper describes a visual QA framework based on ontolgoies, that relies on two main components: question analysis component and answer extraction component. Our goal consists on performing an efficient question answering by: (1) Improving the representation of the questions structure using question ontology and typed attributed graphs, (2) Improving the results of question reformulation using domain ontologies and lexicosyntactic patterns (3) Extracting answers based on the question graph, lexico-syntactic patterns and score computation and (4)Offering a visual representation of the graphs and ontologies. Our framework has been implemented and evaluated.

Ontology-Based Question Analysis Method

Question analysis is a central component of Question Answering systems. In this paper we propose a new method for question analysis based on ontologies QAnalOnto. QAnalOnto relies on four main components: 1 Lexical and syntactic analysis, 2 Question graph construction, 3 Query reformulation and 4 Search for similar questions. Our contribution consists on the representation of generic structures of questions and results by using typed attributed graphs and on the integration of domain ontologies and lexico-syntactic patterns for query reformulation. Some preliminary tests have shown that the proposed method improves the quality of the retrieved documents and the search of previous similar questions.

Semantic Web Search System Founded on Case-Based Reasoning and Ontology Learning

With the continuous growth of data volume on the Web, the search for information has become a challenging task. Ontologies are used to improve the accuracy of information retrieval from the web by incorporating a degree of semantic analysis during the search. However, manual ontology building is time consuming. An automatic approach may aid to solve this problem by analyzing implicitly available knowledge such as the users’ search feedback. In this context, we propose a semantic web search system founded on Case-Based-Reasoning (CBR) and ontology learning that aims to enrich automatically the ontologies by using previous search queries performed by the user. Some experiments and results obtained with the proposed system are also presented, which show an improvement on the precision of the Web search and ontology enrichment.

Fuzzy ontologies for search results diversification: application to medical data

Fuzzy ontologies offer an efficient representation of uncertain information in natural language and this representation allows a better interpretation of user queries and documents. Integrating fuzzy ontologies in a search results diversification process may improve the quality of returned documents since diversification helps covering the maximum of users needs. In this context, we propose an ontology based diversification approach for search results applied to medical domain. The proposal first analyses the query in order to extract medical concepts. A contextual ontology fuzzification is then applied in order to offer an understanding of the users information needs and finally a fuzzy search result diversification is performed in order to improve the ranking quality of returned documents. We perform a thorough experimental evaluation of our proposal with CLEF e-health 2016 topics. Evaluation results show a major improvement in precision and ranking.