Nathalie Aussenac-Gilles

Revisiting Ontology Design: A Methodology Based on Corpus Analysis

We promote a new approach for knowledge modelling based on knowledge elicitation from technical documents. It benefits of the increasing amount of available electronic texts and of the maturity of natural language processing tools. The approach defines a framework where the knowledge engineer selects the appropriate tools, combines their use and interprets their results to build up a domain model. The paper presents the method and reports an on-going application to design an ontology of knowledge engineering tools in French.

Semantic cores for representing documents in IR

This paper deals with the use of ontologies for Information Retrieval. Roughly, the proposed approach consists in identifying important concepts in documents using two criterions, co-occurrence and semantic relatedness and then disambiguating them via an external general purpose ontology, namely WordNet. Matching the ontology and a document results in a set of scored concept-senses (nodes) with weighted links. This representation, called semantic core of a document best reveals the semantic content of the document. We regard our approach, of which the first evaluation results are encouraging, as a short but strong step toward the long term goal of Intelligent Indexing and Semantic Retrieval.

Conceptual indexing based on document content representation

This paper addresses an important problem related to the use of semantics in IR. It concerns the representation of document semantics and its proper use in retrieval. The approach we propose aims at representing the content of the document by the best semantic network called document semantic core in two main steps. During the first step concepts (words and phrases) are extracted from a document, driven by an external general-purpose ontology, namely WordNet. The second step a global disambiguation of the extracted concepts regarding to the document leads to build the best semantic network. Thus, the selected concepts represent the nodes of the semantic network whereas similarity measure values between connected nodes weight the links. The resulting scored concepts are used for the document conceptual indexing in Information Retrieval.

Designing and evaluating patterns for ontology enrichment from texts

Pattern-based approaches for knowledge identification in texts assume that linguistic regularities always characterise the same kind of knowledge, such as semantic relations. We report the experimental evaluation of a large set of patterns using an ontology enrichment tool: Cameleon. Results underline the strong corpus influence on the patterns efficiency and on their meaning. This influence confirms two of the hypotheses that motivated to define Cameleon as a support used in a supervised process: (1) patterns and relations must be adapted to each project; (2) human interpretation is required to decide how to report in the ontology the pieces of knowledge identified with patterns.

A study on LIWC categories for opinion mining in Spanish reviews

With the exponential growth of social media, that is, blogs and social networks, organizations and individual persons are increasingly using the number of reviews of these media for decision-making about a product or service. Opinion mining detects whether the emotions of an opinion expressed by a user on Web platforms in natural language are positive or negative. This paper presents extensive experiments to study the effectiveness of the classification of Spanish opinions in five categories: highly positive, highly negative, positive, negative and neutral, using the combination of the psychological and linguistic features of LIWC (Linguistic Inquiry and Word Count). LIWC is a text analysis software that enables the extraction of different psychological and linguistic features from natural language text. For this study, two corpora have been used, one about movies and one about technological products. Furthermore, we conducted a comparative assessment of the performance of various classification techniques, J48, SMO and BayesNet, using precision, recall and F-measure metrics. The findings revealed that the positive and negative categories provide better results than the other categories. Finally, experiments on both corpora indicated that SMO produces better results than BayesNet and J48 algorithms, obtaining an F-measure of 90.4 and 87.2% in each domain.

Towards a Contextual Pragmatic Model to Detect Irony in Tweets

This paper proposes an approach to capture the pragmatic context needed to infer irony in tweets. We aim to test the validity of two main hypotheses: (1) the presence of negations, as an internal propriety of an utterance, can help to detect the disparity between the literal and the intended meaning of an utterance, (2) a tweet containing an asserted fact of the form Not(P1) is ironic if and only if one can assess the absurdity of P1. Our first results are encouraging and show that deriving a pragmatic contextual model is feasible.

Evaluation of the OQuaRE framework for ontology quality

The increasing importance of ontologies has resulted in the development of a large number of ontologies in both coordinated and non-coordinated efforts. The number and complexity of such ontologies make hard to ontology and tool developers to select which ontologies to use and reuse. So far, there are no mechanism for making such decisions in an informed manner. Consequently, methods for evaluating ontology quality are required. OQuaRE is a method for ontology quality evaluation which adapts the SQuaRE standard for software product quality to ontologies. OQuaRE has been applied to identify the strengths and weaknesses of different ontologies but, so far, this framework has not been evaluated itself. Therefore, in this paper we present the evaluation of OQuaRE, performed by an international panel of experts in ontology engineering. The results include the positive and negative aspects of the current version of OQuaRE, the completeness and utility of the quality metrics included in OQuaRE and the comparison between the results of the manual evaluations done by the experts and the ones obtained by a software implementation of OQuaRE.

A Comparison of Dimensionality Reduction Techniques for Web Structure Mining

In many domains, dimensionality reduction techniques have been shown to be very effective for elucidating the underlying semantics of data. Thus, in this paper we investigate the use of various dimensionality reduction techniques (DRTs) to extract the implicit structures hidden in the Web hyperlink connectivity. We apply and compare four DRTs, namely, principal component analysis (PCA), non-negative matrix factorization (NMF), independent component analysis (ICA) and random projection (RP). Experiments conducted on three datasets allow us to assert the following: NMF outperforms PCA and ICA in terms of stability and interpretability of the discovered structures; the well- known WebKb dataset used in a large number of works about the analysis of the hyperlink connectivity seems to be not adapted for this task and we suggest rather to use the recent Wikipedia dataset which is better suited.

Evaluating a conceptual indexing method by utilizing wordnet

This paper describes our participation to the English Girt Task of CLEF 2005 Campaign. A method for conceptual indexing based on WordNet is used. Both documents and queries are mapped onto WordNet. Identified concepts belonging to WordNet synsets are extracted from documents and queries and those having a single sense are expanded. All runs are carried out using a conceptual indexing approach. Results prove a primacy of using queries from the title field of the topics and a slight gain of using stemming compared to the non stemming cases. ACM Categories and Subject Descriptors H3.3 [Information Storage And Retrieval]: Information Search and Retrieval; H.3.1 [Content Analysis and Indexing] – Search process, Retrieval models. General Terms: Algorithms, Experimentation.

DYNAMO-MAS: a Multi-Agent System for Ontology Evolution from Text

Manual ontology development and evolution are complex and time-consuming tasks, even when textual documents are used as knowledge sources in addition to human expertise or existing ontologies. Processing natural language in text produces huge amounts of linguistic data that need to be filtered out and structured. To support both of these tasks, we have developed DYNAMO-MAS, an interactive tool based on an adaptive multi-agent system (adaptive MAS or AMAS) that builds and evolves ontologies from text. DYNAMO-MAS is a partner system to build ontologies; the ontologist interacts with the system to validate or modify its outputs. This paper presents the architecture of DYNAMO-MAS, its operating principles and its evaluation on three case studies.