Section: Other Grants and Activities
We take part in the Knowledge Web Network of Excellence. This year, we took part in the workpackages WP2.2 Heterogeneity where we worked on ontology alignment, WP2.3 Dynamics where we worked on ontology evolution, WP3.2 and WP3.3 on Education.
Sylvain Dehors took part in the Knowledge Web general assembly in Trento in January 2006.
SeaLife is a 3 year-long STREPS project, coordinated by Dresden University, with Edinburgh University, London College, Manchester University, Scionics as other partners; it started on April 2006.
The objective of SeaLife is the design and realisation of a semantic Grid browser for the Life Sciences, which will link the existing Web to the currently emerging eScience infrastructure  . The SeaLife browser will allow users to automatically link a host of Web servers and WebGrid services to the Web content they are visiting. This will be accomplished using eScience's growing number of WebGrid Services and its XML-based standards and ontologies. The browser will identify terms in the pages being browsed through the background knowledge held in ontologies. Through the use of Semantic Hyperlinks, which link identified ontology terms to servers and services, the SeaLife browser will offer a new dimension of context-based information integration.
This Sealife browser will be demonstrated within three application scenarios in evidence-based medicine, literature and patent mining, and molecular biology, all relating to the study of infectious diseases. The three applications vertically integrate the molecule/cell, the tissue/organ and the patient/population level by covering the analysis of high-throughput screening data for endocytosis (the molecular entry pathway into the cell), the expression of proteins in the spatial context of tissue and organs, and a high-level library on infectious diseases designed for clinicians and their patients.
In this project we take part in 6 among the 7 work packages and we are coordinator of the text mining and natural language processing work package. This year, our main contributions in this project are:
Improvement of our method of annotation generation from text and integration of external sources (Uniprot) to enrich annotations;
State of the art:
Study of different Sealife use cases to define our collaboration with each partner,
Participation in the writing of deliverable A2.D1  ,
Participation in the Sealife kick-off meeting, 10-11 April 2006, Dresden, Germany,
Organization of the Sealife meeting, 13-14 November 2006, Sophia Antipolis, France.
Olivier Corby, Rose Dieng-Kuntz, Fabien Gandon and Khaled Khelif attended the SeaLife kick-off meeting in Dresden on April 10-11, 2006 and the SeaLife second general meeting organized by Acacia at INRIA Sophia Antipolis on November .
Palette is a 3-year long integrated project, coordinated by ERCIM and EPFL, with partners such as Univ. Fribourg, CTI (Greece), Centre de Recherche Public Henri Tudor, University Abou Bekr Belkaid, University of Liège, EM Lyon, Groupe d'Analyse et de Théorie Economique (GATE CNRS), Center for Study of Education and Training (CSET), ePrep, Nisai, MindOnSite - Integral Coaching SA, LICEF Téluq, INRIA Rhône-Alpes.
As scientific objectives, Palette aims at offering information services, knowledge management services (based on an ontology dedicated to communities of practice) and mediation services for communities of practice (CoPs). CoPs are groups of people who share a passion for something that they know how to do, and who interact regularly in order to learn how to do it better. CoPs can be found within businesses, across business units or across company boundaries, still they differ from business or functional units, from teams and networks: people belong to CoPs at the same time as they belong to other organizational structures. CoPs can be considered as a means by which knowledge is owned in practice. Indeed, such groups allow the functions of creation, accumulation and diffusion of tacit and explicit knowledge in organizations.
As applicative objectives, Palette services will be adapted to CoPs in management, in engineering and in educational field.
Acacia is leader of the WP3 aimed at designing ontologies and ontology-based Services for Knowledge Management in Communities of Practice. In the context of the Palette project, knowledge management services are to be provided to CoPs to facilitate the efficient and effective management of their knowledge resources.
Construction of a generic ontology
Knowledge management services will rely on CoP-dependent ontologies to describe and represent the CoPs universe. The first step toward the development of these ontologies was to identify, specify, and organize generic concepts related to the notion of ``community'', forming a generic meta-model that will guide us in the CoP-dependent ontologies development. The main concepts identified enable us to describe a CoP in terms of: its actors; the competencies handled in it; the collaboration between the actors to enhance their learning; the activities performed by the CoP's members; and the lessons learnt from the members' experiences, and through which the CoP's best practices can be identified.
Our contributions to the generic ontology construction task (in particular, the model for Community, Actor, Competency and Lessons-learnt) are fully described in the Palette deliverable D.KNO.01  , and synthesized in the PAKM 2006 conference paper  .
Construction of CoP-dependent ontologies
We are currently developing the CoP-dependent ontologies according to a methodology we elaborated for the Palette project:
Data collection and scenarios analysis: for the building of a contextualized lexicon that contains the candidate-terms as well as their respective contexts of appearance.
Terminological study: the analysis of the contextualized lexicon should lead to the identification of the vocabulary, which is the set of terms of the lexicon, each associated to its definition. The definition of a term is deduced from the information provided by the contextualized lexicon.
Ontology structuring: which consists in identifying the terminological concepts and relations and organizing them, through the building of a concept-hierarchy and a relation-hierarchy  .
Ontology formalization: in which we determine the ontology's concepts and semantic relations (to which we attach their respective domain and range), translate them into RDFS formal ontology.
Design and development of Knowledge Management Services
The developed ontologies will be used to develop knowledge management services (KM services) for CoPs. This task is the main objective of the third work-package (WP3) of the Palette project and implies a strong interaction with other WPs, mainly:
WP1 to gather information about CoPs;
WP5 concerning integration issue, and the links between KM services and other proposed services in Palette (Information and Mediation).
To achieve this task we first specified the KM services (see deliverable D.KNO.03  ). We proposed an architecture of KM services based on a service oriented-architecture (SOA) containing two kinds of services (simple and complex ones) and a repository describing them. We also described a mechanism of services composition that enables us to build specific services for the Palette CoPs, using the simple KM services as building blocks. These building blocks realize the basic tasks of knowledge management, including: Knowledge creation, Ontology creation, Annotation, Collaborative knowledge creation, Knowledge retrieval and dissemination, Knowledge visualization and presentation, Knowledge evolution, Knowledge evaluation.
KM services, Mediation services and Information services interoperability
A critical issue in KM services development is their interoperability and integration with other Palette's services and with CoPs existing tools (the associated work to this issue holds in the WP1 and WP5). To deal with this issue the Palette philosophy advocates the use of W3C standards. We propose to use RDF/S for knowledge representation, and WSDL, SAWSDL, for services. The presentation of these standards and the justification of our choices are described in the deliverable D.IMP.01  . On the other hand, we contributed to the task of bridging the gap between the developers and the end-users, by clearly defining the notions of scenarios and use-cases, and we tried to formalize some of them describing the CoPs activities and the place of the developed services and tools in these activities. This task of clarifying the notion of scenario, and of elaborating a shared representation of this design artefact, was also undertaken in the context of the WP1, to which we also contributed (see the deliverable D.PAR.02  ).
SevenPro (Semantic Virtual Engineering Environment for Product Design) is a European STREPS project. The SevenPro project develops technologies and tools supporting deep mining of product engineering knowledge from multimedia repositories and enabling semantically enhanced 3D virtual reality (VR) interaction with product knowledge in integrated engineering environments. It aims at helping an engineer to design new objects by providing a 3D viewing of the object designed, informations on each part of the object (suggestions of other objects with similar or close properties could be performed) and information about repetitive design processes.
SevenPro is coordinated by Semantic Systems (Spain), and involves partners from industrial and academic areas. SevenPro project is carried out during 34 months, starting from January 2006, by a consortium composed of partners from five different EU countries:
Three centres of excellence with proven expertise in knowledge extraction, data mining and VR technologies: ACACIA (France), Technical university of Prague (Czech Republic), Fraunhofer institute (Germany). One major objective for these centres is to perform research, technology, and development (RTD) and disseminate results;
Two IT companies (high-tech small and medium enterprises: SMEs) specialised in Semantic and Virtual Reality for engineering technologies: Semantic Systems (Spain), and LivingSolids (Germany). The two companies are motivated for exploiting commercially the results, which will perform RTD and take the lead in deploying the technology in the testbeds and later on in the market;
Two users: one SME manufacturer of metal castings: Estanda (Spain) and one big engineering company: ItalDesign (Italy), both having their own engineering teams involved in the project. They will provide requirements and suited testbeds for the technology used in the SevenPro project.
The components and the whole engineering environment that are to be developed will be integrated and tested continuously in an iterative and incremental strategy.
Acacia is leader of WP02 on Knowledge Engineering and WP04 on Semantic Annotation of Corporate Repositories.
This year, we participated in two main WPs. Hacène Cherfi is WP leader in the Knowledge engineering (WP02) with formalization in RDF/S, and when necessary OWL, and Task leader in the WP03 for text annotation prototype.
WP02 – Knowledge Engineering
We designed the Generic Layer ontology for product design on the basis of Use Cases coming from real-life needs and difficulties facing the users during design processes. Use cases were identified by a questionnaire that we have elaborated and distributed to the users, and to all partners in SevenPro. On April, the questionnaire was first presented, and explained during a general meeting, held in INRIA Sophia Antipolis: the aims of the identification phase for the user needs was thus set. The questionnaire was refined with user feedback during Prague general meeting on June. We have also supplied the project with a server for the knowledge entered by users that enables access and queries to a knowledge base. The server was set up, and we aim at using it effectively to store our future front-end applications and our designed ontologies. A significant effort was made, by ACACIA, to help partners to acquire Semantic Web technology and languages. More precisely, we had some requirements on RDF/S expressiveness to formalize some situations that frequently occur during an engineering product design. These situations are very innovative and might raise some Use Cases that supply the W3C with expressiveness needs. An investigation on the usefulness of the solutions that we have proposed will, probably, be subject to a scientific publication in the near future.
WP03 – Prototyping and user feedback
Part of the integration requires development of front-end tools, since components have to be integrated in an engineering environment. ACACIA is in charge of the current development of a text annotation prototype. We have presented, and agreed on, our prototype architecture during the Magdeburg (Germany) general meeting.
The approach is to extract the meaning from the documents themselves; in addition to use of other sources of semantic information. Ontologies come into play at this point, as they constitute a reference knowledge base for the corresponding annotations automatically added to the data. Indeed, the documents are so numerous that we cannot manually supply authors with metadata that suit their meaning and describe their published documents. The annotations are intended to be transparent to the user and rather used by software deployed on the Web (e.g., search engines, crawlers, Web services..).
In this context, natural language processing (NLP) techniques are useful and appropriate to analyse texts, and capture their meanings. Different off-the-shelf modules from existing NLP tools have to be gathered and arranged: (i) text cleaning to manage different text formats (MS Word, OpenOffice, PDF..); (ii) sentence splitting and chunking; (iii) lemmatisation, (iv) term and relation identification; (v) link to the concepts of an ontology. The testbed consists in real-world multilingual corpora. These corpora come from real-world user documentation. The prototype is intended to give first capabilities of NLP techniques to fulfil the knowledge identification within the user documents.
Olivier Corby and Rose Dieng-Kuntz attended the kick-off meeting in Bilbao, Spain, on January 16-17, 2006. The second general meeting was organized by Acacia at INRIA Sophia Antipolis and attended by Olivier Corby, Rose Dieng-Kuntz, Fabien Gandon and Alain Giboin on April 5-6, 2006. Fabien Gandon attended the third general meeting in Prague on June 13-15. Hacène Cherfi, Olivier Corby and Rose Dieng-Kuntz attended the fourth meeting in Magdeburg on September 25-26.