Section: New Results
Keywords : Corporate Memory, Corporate Semantic Web, Cognitive Sciences, Knowledge Acquisition, Knowledge Management, Knowledge Engineering, Ontology, Assistance to the User, Cognitive Psychology, Communication, Co-operation, Human-machine interaction.
Support to Modeling and Building of a Corporate Semantic Web
The objective of this action is to propose methodological and software support for the construction of a corporate memory, through a user-centered approach. We study in particular the construction of a corporate semantic Web and the construction of ontologies and annotations from human and textual sources of expertise or from databases. Moreover, we study how to handle multiple viewpoints or multiple ontologies and how to take into account the life cycle and the evolution of a corporate semantic web. We also study e-Learning as a specific scenario of knowledge management. Recently, we started to study how to tackle the context and privacy of the (possibly mobile) user and how to develop and use ``Corporate Semantic Web Services''.
Designing User-Adapted Semantic Web Applications
The goals are: (1) To propose methods and models to help the user-oriented design of Semantic Web applications; in particular, importing and adapting methods from the Human-Computer Interaction (or software ergonomics) and CSCW communities to the Semantic Web community, esp. the ontology engineering community. (2) To study practices of users and communities of users to inform the design of Semantic Web applications.
Adaptation of scenario-based methods to the design and evaluation of Semantic Web applications.
We continued our work on the adaptation of scenario-based methods to the design and evaluation of Semantic Web applications and of their underlying ontologies (Context: European projects Palette and SevenPro; RNTL project e-WoK HUB). In particular: To identify the various forms under which the partners of a project (user, requirements analyst, designer, developer, evaluator, etc.) represent scenarios and how these partners can pass from a representation to another so that they can understand each other, we elaborated a questionnaire based on the Crews model of a scenario [cf. Palette deliverable D.PAR.02]. As part of a research action aimed at designing an editor of multimedia scenarios for ontology-based applications, we specified interface components of the OntologyCreator tool, allowing performing the initial non formal steps of the ontology construction process.
Adaptation of the Alan Cooper's persona technique to the design of intranets.
The persona technique, a part of the Interaction Design methodology of Cooper, is used to model users of the systems to be designed. A persona is a fictional person created to represent a particular class of real users or system participants. A persona is an archetypal user, who resembles several people, but does not exactly match any one of them. Personas embody the key trends elicited from interviews of users, in terms of distinct sets of behavior patterns and goals  . We used and adapted the persona technique to inform the redesign of an intranet. Intranet personas were identified, and served to elicit recommendations for improving the intranet  . (Context: Project UsableIntranet2.)
Extension of the Alan Cooper's persona technique to represent communities.
The persona technique is intended to represent mainly individuals. However, when we are to design tools aimed at groups of people or communities, we need to represent such groups or communities. To extend the persona technique to the design of systems, we started a study of collective personas, or communities, which can help model particular and concrete groups of people or communities  . (Context: Project UsableIntranet2.)
Study of users' annotation practices for improving intranet search.
Intranets are a place where we can observe users' annotations practices, and the usage of tools supporting such practices. We performed a usage study of the Important Links functionality of an intranet, leading to a set of recommendations and mock-ups for improving the functionality  . The personas elaborated in  were successfully reused for this improvement task. (Context: Project UsableIntranet2.)
Study of users' search practices for adapting search algorithms to users' needs.
In order to adapt Web search algorithms to users' needs, we designed an experiment aiming at identifying the strategies used by scientific researchers to retrieve documents, and at comparing these search strategies and their results with the search strategies and results of the Latent Semantic Analysis (LSA) method and of the Corese method. This experiment is under development. (Context: Action COLOR Edccaeteras.)
Corporate memory and semantic web for the Transcriptome Analysis
The study of gene expression has been greatly facilitated by biochip technology. Biochips can assess tens of thousands of genes simultaneously and lead to a huge amount of information: for example, information about the roles played by particular genes in drug sensitivity and the effects of drugs on gene expression. In the framework of a collaborative project with biologists working on biochip experiments at IPMC (Institut de Pharmacologie Moléculaire et Cellulaire), we aim at assisting them in their experiments and facilitating their validation and interpretation of obtained results. Our objective is to propose methodological and software support for capitalization and valorisation of knowledge resulting from experiments and techniques to preserve and reuse data (structured documents, semantic information retrieval). We rely on the techniques of semantic web (semantic annotations, ontology...) and knowledge engineering. After studying biologists' needs, we proposed to build an experiment memory and to materialize it through a corporate semantic web. The main modules of this memory are:
MeatOnto, a modular ontology composed of 3 subontologies: (1) UMLS (Unified Medical Language System) to describe the biomedical domain; (2) MGED (Microarray Gene Expression Data) covering the technical aspects of the biochip experiments and (3) DocOnto which describes (a) metadata about scientific articles and about annotations, (b) the structure of articles and links documents to UMLS concepts.
MeatAnnot, a system for the automatic generation of ontology-based semantic annotations: starting from a scientific article in biology, it allows us to generate a structured semantic annotation, based on a domain ontology, and describing the semantic contents of this text. MeatAnnot relies on several NLP techniques (e.g. modules of GATE (General Architecture for Text Engineering), RASP (Robust Accurate Statistical Parsing) parser and a relation extraction grammar we wrote in JAPE); it extracts information from text, instantiates concepts and relationships of UMLS ontology and generates RDF annotations for the document.
MeatSearch, the search module based on Corese: it enables biologists to use annotations. By using the query and rule languages of Corese, this system allows us to perform reasoning on the annotation base for retrieving relevant information.
The annotations generated by MeatAnnot were validated by biologists and obtained good scores: (82% of precision, 62% of recall and 96% of usefulness).
Our method can be generalized to any life science domain (e.g. chemistry, physics) having similar needs of support to validation and interpretation of experimental results.
The originality of this work consists of (a) the integration of metadata on annotations which gives new ways of reasoning and more information on the annotation base, (b) the use of several technologies (such as NLP, Ontologies, Semantic annotations, Corese) to build a real-world Corporate Semantic Web Application.
Construction of a multi-point of view Semantic Web
This work is carried out in the context of Thanh-Le Bach's PhD  . The objective of this thesis is to allow the construction and the exploitation of a semantic Web in a heterogeneous organization comprising various sources of knowledge and various user categories, without eliminating heterogeneity but by making heterogeneity and consensus cohabit in the whole organization.
In the previous years, we had proposed three algorithms (ASCO1, ASCO2 and ASCO3) allowing us to align already existing ontologies represented in different ontology languages recommended by W3C for the semantic Web: RDF(S) for ASCO1 and OWL for ASCO2 and ASCO3. This year, we evaluated these algorithms through alignment tool evaluation campaigns, in particular I3CON (the Information Interpretation and Integration Conference) http://www.atl.external.lmco.com/projects/ontology/i3con.html and OAEI (Ontology Alignment Evaluation Initiative).
We also studied the problem of building new ontologies in a heterogeneous organization by taking into account different viewpoints, different terminologies of various users, groups or even communities in the organization. Such ontology, called multi-viewpoint ontology, enables to have both heterogeneity and consensus co-exist in a heterogeneous organization. We proposed a multi-viewpoint knowledge representation model, called MVP, and a multi-viewpoint ontology language, called MVP-OWL, which is an extension of the ontology language OWL, to allow the construction and the exploitation of the multi-viewpoint ontology in a corporate semantic Web.
Management of Corporate Semantic Web Evolution
This work is being carried out within the framework of Phuc-Hiep Luong's PhD, that, based on the analysis of the life cycle of a corporate semantic web (CSW), aims at solving some problems related to its evolution: evolution of each component (resources, ontologies and semantic annotations) as well as evolution of relations among these components.
In the dynamic world, changes in business, technologies and process… of organisations often result in need of evolution of the CSW. This evolution may contain changes in components of a CSW which might affect the consistency in one part or in overall system. Within the scope of research, we have been concentrating on the ontology evolution, its influence on semantic annotations expressed with the vocabulary provided by the underlying ontology and the evolution of these semantic annotations.
We have made a review of some existing work on schema evolution in databases and in knowledge base systems. Several main approaches of the change management for distributed ontologies, existing methods and tools for ontology evolution and the relations between ontologies and semantic annotations have been reviewed as well. We have identified two essential approaches to ontology versioning  and to ontology evolution  . These two main approaches are compared in order to improve their weak points for our current research experiences.
We have made some propositions for managing evolution and especially for solving inconsistency resolution between semantic annotation and ontology when the ontology is modified. We have proposed an evolution management system CoSWEM  that aims at managing the evolution of each component, the evolutionary relation among components and the propagation of the ontology changes to the semantic annotations depending on this ontology. In this system, a rule-based approach is implemented in order to detect inconsistent annotations and to guide the process of solving these inconsistencies by applying correction rules and resolution procedures.
We have also constructed a model of consistency which includes some defined invariants and a mechanism of verification. Based on these consistency constraints, we proposed some inconsistency detection rules enabling to solve the detected inconsistencies of an annotation when its ontology changes, we have established all the possible solutions for each ontology change operation that might have an influence on the annotation consistency. These possible solutions have been called evolution strategies allowing users to select an appropriate way to repair inconsistent semantic annotations.
As further work, we will continue to refine and to formulate the consistency constraints, the inconsistency detection and correction rules as well as the evolution strategies for inconsistent semantic annotations. We will also work on a real scenario with evolving ontology based on RDF(S) and semantic annotations described in RDF language. We will try to formulate invariants of consistency and propagation rules with the help of rules and querying language of Corese semantic search engine.
Semantic Web for E-Learning
This work takes place in the framework of Sylvain Dehors's PhD.
This work explores the potential applications of Semantic Web and knowledge management technologies for e-learning systems. It focuses on e-learning applications for accessing courses and particularly addresses the difficult problem of reusing existing resources in such e-learning systems.
Within this scope, we propose a methodology and an associated system (QBLS) inspired by knowledge management practices and relying on cutting-edge Semantic Web tools and formalisms. The proposal particularly develops three aspects: Reusing existing pedagogical resources through a process called semantization. In this process, pedagogical content is semi-automatically annotated with ontological concepts, from several points of view (domain, pedagogy, structure, etc.). This process aims at contextualizing the material with regard to the teacher's conceptualization and pedagogical strategy. It specifically offers answers to the complexity and difficulty of annotation by a human  ,  ,  .
Exploiting annotated pedagogical resources in dynamic web interfaces.
We show how Semantic Web technologies efficiently perform useful inferences based on ontological knowledge to personalize and adapt courses. The interest of actual Semantic Web standards (OWL, RDF, SPARQL) is highlighted in a practical perspective, relying on the deployment of semantic search engine Corese.
Tracking learner's activity and analysis of this activity.
We explore the potential of graphics-based visualization and inferences to manually or automatically analyze activity traces on a semantized course. The proposed tools and methods for analysis take advantage of the semantics and underlying Semantic Web framework. The theoretical proposal is supported by our implementation of the Question Based Learning System where learning resources available on the web are reused, annotated and exploited in real world experiments using Semantic Web technologies. The practical examples mentioned in this work are illustrated by two experiments conducted at the EPU of Nice - Sophia Antipolis.