Thierry Despeyroux | Searcharxiv

arXiv2011

Multi-representation d'une ontologie : OWL, bases de donnees, systèmes de types et d'objets

Due to the emergence of the semantic Web and the increasing need to formalize human knowledge, ontologie engineering is now an important activity. But is this activity very different from other ones like software engineering, for example ? In this paper, we investigate analogies between ontologies on one hand, types, objects and data bases on the other one, taking into account the notion of evolution of an ontology. We represent a unique ontology using different paradigms, and observe that the distance between these different concepts is small. We deduce from this constatation that ontologies and more specifically ontology description languages can take advantage of beeing fertilizated with some other computer science domains and inherit important characteristics as modularity, for example.

cs.IR↗

arXiv2007

Extraction d'entités dans des collections évolutives

The goal of our work is to use a set of reports and extract named entities, in our case the names of Industrial or Academic partners. Starting with an initial list of entities, we use a first set of documents to identify syntactic patterns that are then validated in a supervised learning phase on a set of annotated documents. The complete collection is then explored. This approach is similar to the ones used in data extraction from semi-structured documents (wrappers) and do not need any linguistic resources neither a large set for training. As our collection of documents would evolve over years, we hope that the performance of the extraction would improve with the increased size of the training set.

cs.IR↗

arXiv2006

Developing efficient parsers in Prolog: the CLF manual (v1.0)

This document describes a couple of tools that help to quickly design and develop computer (formalized) languages. The first one use Flex to perform lexical analysis and the second is an extention of Prolog DCGs to perfom syntactical analysis. Initially designed as a new component for the Centaur system, these tools are now available independently and can be used to construct efficient Prolog parsers that can be integrated in Prolog or heterogeneous systems. This is the initial version of the CLF documentation. Updated version will be available online when necessary.

cs.SE↗

arXiv2005

Practical Semantic Analysis of Web Sites and Documents

As Web sites are now ordinary products, it is necessary to explicit the notion of quality of a Web site. The quality of a site may be linked to the easiness of accessibility and also to other criteria such as the fact that the site is up to date and coherent. This last quality is difficult to insure because sites may be updated very frequently, may have many authors, may be partially generated and in this context proof-reading is very difficult. The same piece of information may be found in different occurrences, but also in data or meta-data, leading to the need for consistency checking. In this paper we make a parallel between programs and Web sites. We present some examples of semantic constraints that one would like to specify (constraints between the meaning of categories and sub-categories in a thematic directory, consistency between the organization chart and the rest of the site in an academic site). We present quickly the Natural Semantics, a way to specify the semantics of programming languages that inspires our works. Then we propose a specification language for semantic constraints in Web sites that, in conjunction with the well known ``make'' program, permits to generate some site verification tools by compiling the specification into Prolog code. We apply our method to a large XML document which is the scientific part of our institute activity report, tracking errors or inconsistencies and also constructing some indicators that can be used by the management of the institute.

cs.IR↗

arXiv2005

Experiments in Clustering Homogeneous XML Documents to Validate an Existing Typology

This paper presents some experiments in clustering homogeneous XMLdocuments to validate an existing classification or more generally anorganisational structure. Our approach integrates techniques for extracting knowledge from documents with unsupervised classification (clustering) of documents. We focus on the feature selection used for representing documents and its impact on the emerging classification. We mix the selection of structured features with fine textual selection based on syntactic characteristics.We illustrate and evaluate this approach with a collection of Inria activity reports for the year 2003. The objective is to cluster projects into larger groups (Themes), based on the keywords or different chapters of these activity reports. We then compare the results of clustering using different feature selections, with the official theme structure used by Inria.

cs.IR↗

arXiv2005

Expériences de classification d'une collection de documents XML de structure homogène

This paper presents some experiments in clustering homogeneous XMLdocuments to validate an existing classification or more generally anorganisational structure. Our approach integrates techniques for extracting knowledge from documents with unsupervised classification (clustering) of documents. We focus on the feature selection used for representing documents and its impact on the emerging classification. We mix the selection of structured features with fine textual selection based on syntactic characteristics.We illustrate and evaluate this approach with a collection of Inria activity reports for the year 2003. The objective is to cluster projects into larger groups (Themes), based on the keywords or different chapters of these activity reports. We then compare the results of clustering using different feature selections, with the official theme structure used by Inria.

cs.IR↗