Papers by Bernard Espinasse
Demonstration of the Localization of Muscarinic Acetylcholine Receptors in the Gastric Mucosa Light and Electron Microscopic Autoradiographic Studies Using 3H Quinuclidinyl Benzilate Light and Electron Microscopic Autoradiographic Studies Using 3 H Quinuclidinyl Benzilate
日本組織細胞化学会, 1984
L'imaginaire et le virtuel au service de la gestion réelle de crise
Understanding, reusing, and maintaining data warehouse resources is a key challenge for data ware... more Understanding, reusing, and maintaining data warehouse resources is a key challenge for data warehouse users. Data warehouses resources are shared by different groups of users. The interpretation of information is subjective, it depends on user knowledge. Thus, a resource, like a data cube, is interpreted differently from a user to another. Unfortunately, misinterpreting data could induce serious problems and conflicts. To guarantee homogenous interpretation of data warehouse resources additional information is necessary. To tackle these challenges we propose to use ontologies to help the users in the exploitation of data warehouses. In this paper we propose an ontology-driven approach that represents data warehouse, dimensions and facts semantically enriched by their equivalent domain concepts and related to final resources provided by this data warehouse.
Relevant information gathering in the Web is a very complex task. The main problem with most info... more Relevant information gathering in the Web is a very complex task. The main problem with most information retrieval approaches is neglecting the context of the pages, mainly because search engines are based on keyword-based indexing. Considering restrained domain, it is possible to take into account of this contextwhat should lead to more relevant information gathering. In this paper, a specific cooperative information gathering approach based on the use of software agents and ontologies is proposed. To operationalise this approach, a generic software architecture, named AGATHE, permitting the development of specific restricted-domain information gathering systems is presented in detail.
Data Warehouses (DW) resources are shared by users’ from different backgrounds (e.g., domain, cul... more Data Warehouses (DW) resources are shared by users’ from different backgrounds (e.g., domain, culture, education, profession). Those resources (e.g., OLAP queries, Excel files) are interpreted differently from a user to another. Unfortunately, misinterpreting data could induce serious problems and conflicts. To guarantee relevant interpretation of resources, additional semantic description of resources concepts is necessary. In this context, we present an Ontology-driven Personalization System (OPS) based on three connected ontologies: domain ontology, DW ontology and resources ontology. OPS return a set of personalized resources search based on users’ domain and his recurring interests. In addition, resources are enhanced with a semantic description provided by the ontologies. This paper focuses on the methodology used to develop connected ontologies used by OPS. Keywords-data warehouse, ontology, personalization, decision support systems, decision making, healthcare institution ma...
Protocoles de coopération pour le réordonnancement d'atelier
In the last decade, Internet is used, among others things, as an educational information source. ... more In the last decade, Internet is used, among others things, as an educational information source. To help in storage, classification and reuse of educational resources appears the concept of Learning Objects (LO) in order to classify educational material, to provide modular units of learning with metadata, and to improve the access and reuse of them. In this work we analyze, on the one hand, the importance of metadata in Learning Objects in order to obtain a personalized recommendation. On the other hand, exploring the state of the art of automatic metadata extraction, we analyze different software systems and we make a comparison of these systems. Finally, we make some conclusions about several lines of possible research work to address the problem of lack of metadata information in LOs.
Analyse multi-agents de la gestion hydraulique de la Camargue : Considération méthodologiques
De l'Evaluation à l'Adaptation d'un Environnement de Jeu Sérieux : Application au Domaine de Gestion de Crise
Rooster RÉSUMÉ. La dématérialisation des processus de recrutement n'a pas fait disparaître toutes... more Rooster RÉSUMÉ. La dématérialisation des processus de recrutement n'a pas fait disparaître toutes les frictions inhérentes à cette activité. La recherche automatisée d'un candidat idéal se heurte toujours à la difficulté à modéliser correctement les besoins exprimés en langage naturel dans une offre d'emploi. Le recrutement d'experts, notamment, est particulièrement difficile. En effet, ces profils concernent une proportion réduite des recrutements et leur prise en charge informatisée nécessite une connaissance précise du secteur d'activité concerné. Dans cet article, nous proposons l'architecture d'un système de recommandation de profils experts dans l'industrie des procédés afin d'assister ce type de recrutements. ABSTRACT. The digititalization of recruitment processes did not smooth out all frictions related to this activity. Automatically searching for an ideal candidate is still hindered by one major issue: adequately modeling the needs expressed in natural language in a job offer. This problem is even more obvious when one considers the use case of recruiting expert profiles. Indeed, these profiles represent a minority of the recruitments. To assist a recruitment process concerning these profiles, an automated system requires an accurate knowledge of their activity sector. In this article we introduce the architecture of an expert profile recommender system applied to the process industry sector. MOTS-CLÉS : Système de recommandation de poste, Recherche d'expert, Web sémantique.
Unsupervised extractive multi-document summarization method based on transfer learning from BERT multi-task fine-tuning
Text representation is a fundamental cornerstone that impacts the effectiveness of several text s... more Text representation is a fundamental cornerstone that impacts the effectiveness of several text summarization methods. Transfer learning using pre-trained word embedding models has shown promising ...
This paper concerns supervised classification of text. Rocchio, the method we choose for its effi... more This paper concerns supervised classification of text. Rocchio, the method we choose for its efficiency and extensibility, is tested on three reference corpora "20NewsGroups", "OHSUMED" and "Reuters", using several similarity measures. Analyzing statistical results, many limitations are identified and discussed. In order to overcome these limitations, this paper presents two main solutions: first constituting Rocchio-based classifier committees, and then using semantic resources (ontologies) in order to take meaning into consideration during text classification. These two approaches can be combined in a Rocchio-based semantic classifier committee.
Recently, Crisis Management Serious Games (CMSG) have proved their potential for teaching both te... more Recently, Crisis Management Serious Games (CMSG) have proved their potential for teaching both technical and soft skills related to managing crisis in a safe environment while reducing training costs. In order to improve learning outcomes insured by CMSGs, many works focus on their evaluation. Despite its great interest, the learner emotional state is often neglected in the evaluation process. Indeed, negative emotions such as boredom or frustration degrade the learning quality since they frequently conduct to giving up the game. This research addresses this gap by combining gaming and affect aspects under an Educational Data Mining (EDM) approach to improve learning outcomes. Therefore, we propose an EDM-based multimodal method for assessing learners' affective states by classifying data communicated in text messaging and facial expressions. This method is applied to assess learners' engagement during a game-based collaborative evacuation scenario. The obtained assessment r...
In this paper we propose to bring together the concept of Serious Game and Intelligent Tutoring S... more In this paper we propose to bring together the concept of Serious Game and Intelligent Tutoring Systems (ITS) in the context of the SIMFOR project, a serious game for training in crisis management. We discuss the problems and needs of serious games and a overview of existing work. To enhance the learning aspect in serious games, we propose the integration of different ITS modules to the serious game. This integration is realized in the design of a collaborative multi-agent system that represents the different module of an ITS.
Unsupervised query-focused multi-document summarization based on transfer learning from sentence embedding models, BM25 model, and maximal marginal relevance criterion
Extractive query-focused multi-document summarization (QF-MDS) is the process of automatically ge... more Extractive query-focused multi-document summarization (QF-MDS) is the process of automatically generating an informative summary from a collection of documents that answers a pre-given query. Sentence and query representation is a fundamental cornerstone that affects the effectiveness of several QF-MDS methods. Transfer learning using pre-trained word embedding models has shown promising performance in many applications. However, most of these representations do not consider the order and the semantic relationships between words in a sentence, and thus they do not carry the meaning of a full sentence. In this paper, to deal with this issue, we propose to leverage transfer learning from pre-trained sentence embedding models to represent documents’ sentences and users’ queries using embedding vectors that capture the semantic and the syntactic relationships between their constituents (words, phrases). Furthermore, BM25 and semantic similarity function are linearly combined to retrieve...
Supply Restoration in Electric Distribution Networks : A Multi-Agent Approach
Energy Management Systems can benefit from the multi-agent approach at different levels. This app... more Energy Management Systems can benefit from the multi-agent approach at different levels. This approach offers potential to bring efficient solutions to the problem of system supervision and supply restoration in the event of contingencies. The goal of this paper is to present a supply restoration method for electric distribution based on the repair solution philosophy as opposed to the analytical methods traditionally employed to solve this problem. To implement this restoration method, a multiagent model of the network is proposed in order to find, by cooperation between agents, a satisfactory balance between the available supply and its restoration in the network. This research is developed on the basis of a study performed on the operation of a real electricity company. Since the electricity supply is a crucial service for the society, all electricity network equipment is subjected to voltage and current constraints in order to operate within safe and efficient levels. For that p...
Les systèmes de résumé automatique de textes (SRAT) consistent à produire une représentation cond... more Les systèmes de résumé automatique de textes (SRAT) consistent à produire une représentation condensée et pertinente à partir d’un ou de plusieurs documents textuels. La majorité des SRAT sont basés sur des approches extractives. La tendance actuelle consiste à s’orienter vers les approches abstractives. Dans ce contexte, le résumé guidé défini par la campagne d’évaluation internationale TAC (Text Analysis Conference) en 2010, vise à encourager la recherche sur ce type d’approche, en se basant sur des techniques d’analyse en profondeur de textes. Dans ce papier, nous nous penchons sur le résumé automatique guidé de textes. Dans un premier temps, nous définissons les différentes caractéristiques et contraintes liées à cette tâche. Ensuite, nous dressons un état de l’art des principaux systèmes existants en mettant l’accent sur les travaux les plus récents, et en les classifiant selon les approches adoptées, les techniques utilisées, et leurs évaluations sur des corpus de références. ...
Improving Learners’ Assessment and Evaluation in Crisis Management Serious Games: An Emotion-based Educational Data Mining Approach
Entertainment Computing
Abstract For several years, there has been growing interest in the development and use of serious... more Abstract For several years, there has been growing interest in the development and use of serious games to improve individuals’ quality of life and behavior. In particular, Crisis Management Serious Games (CMSG) have shown their potential for teaching people both technical and soft skills related to managing crises in a safe environment while reducing training costs. To improve their effectiveness, several evaluation approaches of CMSGs have been proposed. However, despite its interest, the learner’s emotional state is often neglected. As a result, learners may end up with a deep frustration or boredom. Therefore, we propose an Educational Data Mining (EDM) approach to evaluate learners’ affective states in collaborative CMSGs. This approach is applied to assess learners’ engagement during a game-based evacuation scenario aiming to train and to raise awareness among students of a university on evacuating the present persons during a fire emergency situation. The experimental results confirm the major predictions of learners’ affective dynamics and uncover novel findings namely the presence of transitions from boredom to frustration and from frustration to confusion. Our study shows that combining gaming and emotion aspects under an EDM approach to evaluate CMSGs is interesting since it gives reliable results in a less invasive way.
Introduction au Big Data - Opportunités, stockage et analyse des mégadonnées
Documents numériques Gestion de contenu
L’objet de cet article est de cerner ce terme Big Data ou megadonnees. Dans un premier temps, les... more L’objet de cet article est de cerner ce terme Big Data ou megadonnees. Dans un premier temps, les megadonnees sont caracterisees au travers du modele des 3V etendu au 5V. La problematique des megadonnees est distinguee de celle de l’informatique decisionnelle. Les enjeux economiques et societaux associes aux megadonnees sont abordes en presentant differents exemples d’usage relevant de differents domaines d’activite. Sont ensuite introduites differentes grandes methodes et techniques associees au stockage et a l’exploitation/analyse de ces megadonnees.
Spatial Role Labeling based on Improved Pre-trained Word Embeddings and Transfer Learning
Procedia Computer Science
Abstract In several real-world applications, extracting spatial semantics from text is critical. ... more Abstract In several real-world applications, extracting spatial semantics from text is critical. Spatial Role Labeling (SpRL) introduces a language-independent annotation scheme used in these applications, particularly for reasoning purposes. This paper proposes, first of all, a transfer learning method with a word embeddings-based approach for SpRL. Then, we enhance the word vectors with POS tags and CNN-based character-level representations. Finally, we propose a Residual BiLSTM CRF deep learning model to identify the spatial roles. The experimental results on two datasets: SemEval-2012 and SemEval-2013 Task 3, show that the proposed model outperforms other machine learning approaches.
Uploads
Papers by Bernard Espinasse