REPORT OF THE ECOINFORMATICS TERMINOLOGY MEETING DRAFT 1 10 May 2004 Sponsored by the United National Environment Programme UN Environment House Geneva, Switzerland 15-16 April 2004 Revised 13 May 2004 Table of Contents Executive Summary The purpose of the meeting was to bring together the major providers of terminologies to discuss how new technologies are being applied and how these valuable resources can be integrated using the Web as a platform for sharing. The United Nations Environment Programme (UNEP), the United States Environmental Protection Agency (USEPA), the European Environment Agency (EEA), and the United States Geological Survey (USGS) are part of the Ecoinformatics Initiative core group. With the Ecoinformatics Technology Working Group, UNEP has taken the lead on terminology. The meeting was attended by ____ organizations. Four types of organizations attended -governments, intergovernmental organizations (especially UN agencies), scientific institutions, and corporations and vendors. There were terminology developers (experts on content and structure), IT professionals interested in putting terminologies on the Web, and those interested in multilinguality. During the first day, the participants presented brief descriptions of terminology resources and their interest in providing these resources via the Web. The challenges and opportunities related to multilinguality and different user groups were discussed during general sessions. The participants expressed a great interest in standards, best practices and general tools such as web services and UDDI registries that could further the applicability of knowledge organization systems on the Web. An example of how two knowledge organization systems could be virtually integrated was presented during the meeting. The SWAD-E work on the SKOS Core 1.0 was also presented. The group agreed to investigate several terminology projects related to web services, to create a listserv to promote better sharing of information among those in attendance, and …. WELCOME AND INTRODUCTION Mr. Frits Schlingemann, Director of the Regional Office for Europe, welcomed the group. He briefly described the work of the 200 people who work in this building, including the four convention secretariats. The purpose of the meeting was to bring together the major providers of terminologies to discuss how new technologies are being applied and how these valuable resources can be integrated using the Web as a platform for sharing. The United Nations Environment Programme (UNEP), the United States Environmental Protection Agency (USEPA), the European Environment Agency (EEA), and the United States Geological Survey (USGS) are part of the Ecoinformatics core group. UNEP has taken the lead on the terminology side. It is important not to work in isolation; collaboration is needed with agreed upon roles. The goal of the meeting was to identify specific projects with which to move forward. This was the first UNEP meeting on terminology since the InfoTerra meeting in Madrid in the early 1990s. Four types of organizations attended -- governments, intergovernmental organizations (especially UN agencies), scientific institutions, and corporations and vendors. There were thesaurus developers (experts on content and structure), IT professionals interested in putting terminologies on the Web, and those interested in multilinguality. The first day of the meeting was an opportunity for information exchange. The second day was a more forward looking strategic day. During that time the group looked at overlapping interests and formulated specific projects. Introduction to Ecoinformatics (Bruce Bargemeyer, USEPA and Stefan Jensen, EEA) Ecoinformatics is the application of information science and information technology to the environment. Informatics, especially bioinformatics and healthcare informatics, are well established, but ecoinformatics is relatively new. The Ecoinformatics Initiative leverages the resources of organizations with major health and environmental programs in their missions, including the EEA, USEPA and several US government agencies, plus international organizations such as the UNEP, the JRC, and GBIF. The cooperation draws together and demonstrates principles, techniques, standards and technologies that can form an ecoinformatics marketplace. This is set in the larger cooperative framework of several priority areas including children’s health and environmental indicators. The Ecoinformatics Terminology Working Group intends to cooperate in deploying information systems internationally, share the costs and benefits of the development and publicize the results. Major projects include sharing experiences and results; fostering an ecoinformatics marketplace; cooperating on emerging technologies, and developing key elements for interoperability including terminology. Technologies that might be relevant include computer grids, metadata registries, the semantic web, XML, terminology systems, and ontology agents (from the JRC and the 5th framework). Suddenly the Internet cares about semantics and meaning. Terminology is of great interest and new technologies care about semantics. Data standards are really driven from terminology, since they depend on .access to vocabulary, terminology and standard metadata. How can we point to these resources in new and meaningful ways especially on the web? Terminology is a major area of cooperation. EEA and EPA are utilizing terminologies and various members of the Ecoinformatics Initiative have developed terminology resources. The Cooperative is looking to the broader group for all kinds of resources. SPIRE is a US National Science Foundation semantic prototype in research ecoinformatics based on semantic web technologies in which the USGS and USEPA are involved. The Ecoinformatics Group has several educational, promotional and technology transfer opportunities planned, including a session at the EPA Science Forum in Washington DC, sessions at the EnviroInfo2004, and involvement in an Ecoinformatics “Birds of a Feather” discussion at Digital Government 2004. The Ecoinformatics Initiative web page was developed and is maintained by EEA; it provides a single location for the documents and news about the group. Through this meeting, the group hopes to further its objectives, find additional areas for cooperation and gain new participants and projects where possible. Types and Varieties of Terminology Resources (Gail Hodge, USGS/NBII) There are a wide variety of terminology resources, or knowledge organization systems (KOS). The term knowledge organization system encompasses all types of schemes for organizing information, performing collection management, promoting discovery and retrieval, and promoting knowledge sharing and management [Hodge, 2000]. (A taxonomy of KOSs has been developed by Hill and Hodge [2000].) KOSs range from term lists such as authority files and glossaries to those that express complex term relationships like ontologies, semantic networks and topic maps. KOSs can be characterized by subject, by structure (the number of terms, the breadth (or the number of top terms), the depth of the hierarchies and the types of relationships expressed), and by function (database indexing, cataloging or metadata creation, shelving, etc.) These characteristics may be useful in describing KOSs so that they can be utilized more automatically on the Web. The full spectrum of KOSs should be considered during the conference discussions. Introduction to Web Services (Tim Lynch, USDA) The term “Web services” is given to the networking technologies that allow web servers or web sites to communicate with one another, computer to computer. Web services are supported by a host of technologies including XML, SOAP, and RSS. More can be found at the W3C site at www.w3.org. Web services are used when purchasing something online. During the purchasing process, the web site (via the submit button) processes your credit card and then if the proper information is returned it processes the order. The activity behind the scenes is actually quite complex, perhaps involving a bid process if there are multiple vendors as with some discount sites. Other examples exist in government, education and scientific research. With a web service, one organization can develop an application in its favorite language and technology, but still communicate almost seamlessly with another. The key attribute of a web service is that it is reliable, secure, scaleable, and works across platforms. These attributes also describe the infrastructure that environmental organizations need in order to communicate with one another. A terminology web service would involve two web sites communicating about terminologies. For example, one site provides a multilingual vocabulary and one provides environmental information in only one language. How can the second site offer multilingual searching? Historically, this has been done by installing the thesaurus at the second web site. However, if the thesaurus is being updated frequently, it requires an update and synchronization of the downloaded thesaurus. With web services this problem is eliminated OVERVIEW OF EXISTING ENVIRONMENTAL THESAURUS/TERMINOLOGY INITIATIVES Each terminology developer presented a brief description of the resource and plans for the future. More details are provided in Appendix D. Environmental Applications Reference Thesaurus - EARth (Paolo Plini, CNR) Thesauri, which were primarily developed in library environments, may not satisfy all the requirements of the new information context. CNR’s vision is to develop a new thesaurus model for the environmental domain that combines a stable logical and conceptual base with flexibility to support different applications. It represents a semantic mapping of the environmental domain and different cultural dimensions to ensure applicability for users with different levels of expertise and portability to different technology applications. The EARTh architecture is based on categories with a faceted structure. The classification includes entities, attributes, dynamic aspects (such as processes, conditions, activities), and dimensions (space and time). Node labels help in the location of terms in the system. Thematic vocabularies will be developed for applications by combining the classification scheme and the structure, allowing users to create their own terminology systems. The expansion of relationship types has already begun with a set of subrelationships. CNR is on the verge of providing the data from the thesaurus as a web service to support organizations that would like to include thesaurus facilities but which have no resources to produce their own thesaurus environments. UDK Thesaurus and UMTHES (Wolf-Dieter Batschi, UBA Germany) The UDK and UMTHES are the same thesaurus. UMTHES is used in several databases (ULIDAT, UFORDAT, the German Environmental Information Network (GEIN), etc.) in environmental catalogues, and in new applications such as text analysis software. It contains over 8,900 preferred terms, non-descriptors and geographic terms in German and English with a poly-hierarchical structure. Definitions, scope notes and a classification scheme are provided. In addition to the standard BTs, NTs, RTs, and synonyms, additional structures have been added to support topic maps. TED - Thesaurus on Emergencies and Disasters (Rudolf Legat, UBA Austria) TED is an integrated Internet portal funded by the Austrian Federal Ministry for Education, Science and Culture for emergency and civil protection agencies, candidate countries, local and regional authorities and UNEP.net. The objective is to bridge the gap between different sources and users of the international and national content selected for the information system. The motivation for developing this thesaurus is the lack of terminological knowledge in the field of disaster information. Disasters go beyond geographic and political boundaries. The basis for TED terms are existing thesauri and terminologies in a variety of related areas. While English is a key language, it isn’t sufficient in emergency situations; and the extension of the EU makes the multilinguality even more important. (The source for the German translation was GEMET.) There is an open invitation for countries and organizations to do translation. National Biological Information Infrastructure (NBII) (Lisa Zolly) The NBII is a virtual network of information providers. Currently there are 12 nodes that provide access to data, information or tools for natural resource managers. The nodes have expertise in a particular biological discipline or geographic region of the United States. The NBII has been active in thesaurus development since its beginning when it worked with the California Resources Evaluation System. However, when a broader thesaurus was needed, the NBII partnered with Cambridge Scientific Abstracts to develop the Biocomplexity Thesaurus. CSA integrated five thesauri and added additional terms required for biodiversity creating a thesaurus of 10,000 terms with standard relationships. Coverage includes fisheries and aquatic sciences, life sciences, social sciences and ecotourism. There are proposals to expand the thesaurus to include fire ecology, fire management, forestry and forest management. The thesaurus is proprietary, but it can be used for certain non-profit uses. For example, the SPIRE semantic web project will use the Biocomplexity Thesaurus. The Biocomplexity Thesaurus is used for any resource provided through the NBII metadata clearinghouse and for open GIS and web resources. In the next few weeks, hierarchical and alphabetical views of the thesaurus will be published in addition to the current browse. Web services are being developed to provide thesaurus look up tools on the desktop and the development of mini thesauri on the fly. The hope is to provide users with more precise searching and more relationships between terms. CAB Thesaurus (James Brooks, CAB International) CAB is an international, intergovernmental not-for-profit organization of 40 member governments. The CABI thesaurus for the Applied Life Sciences was originally developed for Agriculture but it has expanded to cover aquaculture, food science, water resources management and invasive species. It contains about 47,000 descriptors and 12,000 entry terms in English, Spanish, Portuguese and Dutch. Systematic revision of certain areas is underway, including the addition of common names for bacteria and viruses, soil types and geographic instances. In addition, a high level ontology will be added. Other developments include greater granularity of relationship types (partitive, generic and instantive, hierarchical, etc.). A contextual role may also be used. CABI is building a new thesaurus management system to support XML output. They are also looking at the SKOS RDF-S format. CABI is participating with multiple stakeholders, including the FAO Agricultural Ontology Service and the Global Forestry Information System. Research projects include text mining with several British projects organizations and some commercial partnerships with publishers and other content partners. In terms of the work of this group, economic and management issues are important considerations. Do we have the right data models and what infrastructure is required? One which is demand led or user driven? Doesn’t think that the work of the British or the International standards organizations will be broad enough for what we need for interoperability Integrated Taxonomic Information System – ITIS (Janet Gomon, Smithsonian Institution, USA) ITIS is a partnership of US federal agencies. It provides an authoritative scientific opinion for the names of organisms and their placement covering four languages (English, Spanish, French and Portuguese). ITIS can serve as an authority file for environmental databases that involve species. It is structured both as a thesaurus and a taxonomy, with parent-child links and synonyms linked to the scientific names. Published references for current usage, author and date of the original publication of the name, jurisdiction (native or introduced), and some data quality indicators are included. The name records come from three sources, the original NODOC file which preceded ITIS, new names from taxonomic experts who provide lists or taxonomic treatments of species, and users of IT IS who send names for inclusion. Smithsonian staff enter the data working with taxonomic experts on issues of placement and current usage. ITIS has agreements with GBIF to provide names to the Catalog of Life Project. XML output is available, but there is currently no support for web services. SNS – The Semantic Network Service (Maria Ruether, UBA-Germany & Thomas Bandholtz, SchlumbergerSema) The SNS is a shared vocabulary of 39,000 environmental terms in German and English, a gazetteer including intersections between nearly 50,000 geographic names in German and administrative names like national parks and catchment areas. SNS has also developed a chronology of approximately 600 historic and contemporary events that have impacted the environment which are classified by terms such as conferences, marine disasters, etc. In order to integrate this work into a topic map, SNS developed a semantic model. The original structure of the components is retained with a topic type descriptor and relationships between the terms described as associative relations. To develop the semantic network SNS established relationships between the events, locations and thesaurus terms to describe what happened. SNS offers three services – findTopics, autoClassify and getPSI. The Developer Toolkit and documentation is available from the SNS web site. Environmental Multilingual Thesaurus of the Spanish Environment Ministry (Carmen Casal Fornos & Arantza López de Sosoaga, MMA) Since 1981-1982 they have been developing the glossary and multilingual thesaurus of territorial planning, urban development and the environment. A total revision of the terms is underway along with translations into other official Spanish languages. The main characteristic is that it is a real documentation tool based on actual usage in indexing and retrieval in libraries and document centers. There are several main subject categories (microthesauri). Hierarchical and alphabetical lists are available for each language. The Institute of Environment Portugal is developing an environmental thesaurus which will be turned over to the national library for integration with a larger thesaurus. The goal is to harmonise across all sectors. Terminology Co-ordination and Harmonisation in the EU in the light of the IATE Project (Dieter Rummel, EU ) There are several translation services of the EU, including those for the Parliament, Council and Commission. Many of these resources are online either on the Web or internal to the organization. Other organizations have glossaries in Word or on card files without any systematic access. The drawback to this situation is that there is no single point of access where a user can retrieve up-to-date information for all institutions. There is limited cooperation among the language services, which means users must look in multiple places, both locally and internationally. It is hard to cooperate and reduce duplicative effort. The Centre was set up in the 1980s when there was limited access to these resources. A feasibility study conducted in 1999 recommended incorporation of existing terminology databases into a single new institutional database with interactive data entry by linguists, integration of terminology into office automation tools, and development of a collaborative structure for data management. The InterAgency (soon to be renamed Interactive) Terminology Exchange (IATE) will be operational later this year. The current version has 1.4 million concepts and 7 million terms. Problems include detecting and dealing with perfect duplicate entries and nearperfect entries and harmonizing the use of standard values for certain fields such as spelling variants. Validating the quality in such a distributed system is also an issue. The system keeps information about “the author” of the term, his expertise, language, etc. Based on these factors it will allow users to validate data. Multiple validation stages may exist. They have tried to keep this validation process open so that a variety of organizations can participate. A special e-mail and internal messaging system are used to communicate within the groups. The workflow is based on Oracle triggers. One of the other goals of IATE is to improve the inter-institutional management of the terminologies. Institutional working groups on user testing and data management have been established. Multilingual Thesaurus for the GeoSciences (Jan Jellema, TNO-NITG, Netherlands) Geology and the environment are becoming more interrelated and in the future the organizations working in these disciplines will need to cooperate more in terminology as well. Geology is made up of numerous disciplines, just like the environment, which means that homographs must be resolved. The tools required to provide robust terminology use and management are the same in the geosciences as in the environment even though the content is different. The current GeoSciences Thesaurus was developed in 1972. It involves about 12 institutes worldwide and is expanding. The thesaurus contains approximately 10,000 terms and 6000 descriptors in 11 languages. The hydrology and building materials sections have definitions. The first trials of displaying Cyrillic and Chinese characters on the web have gone well. There is a desire to improve the thesaurus to allow for repurposing and to give it wider importance. Plans include RDF and XML implementations, more discipline oriented user interfaces, distributed entry by members of the community with institutes performing validation and integration with Internet searching and knowledge management. GEMET and EEA Glossary (Stefan Jensen, EEA) EEA began development of GEMET (General Multilingual Environment Thesaurus) in 1996 through a consortium of partners. When the Topic Centre closed in 2001, GEMET was put on hold, but there were various institutions still working on it. It is available in 90 languages, with broad coverage including administration, social sciences and legislation. GEMET also has aspects of a dictionary, since there are definitions for approximately 80% of the terms. GEMET has been distributed as a standalone PC tool with a browser to more than 500 users. However, it is difficult to determine what people are doing with GEMET. A questionnaire was distributed in 2001, but the return rate was only about 10%. The EEA Glossary is available in 24 European languages. It has about 1100 terms that occur on the EEA web site. About 50% of the terms in the glossary overlap with GEMET. A multilingual web site will be available by the end of the month with content in all the languages. USEPA Terminology Reference Services (Larry Fitzwater, US Environmental Protection Agency) Terminology is key to the successful use and reuse of EPA data and other resources. EPA did not have standard ways of naming elements and there were multiple definitions across EPA, the federal government and the state partners, making it difficult to share and use data, particularly from legacy systems. The Terminology Reference System (TRS) collects terms, definitions, and the context, along with a mapping to the legislation that created the definitions. It is based in part on GEMET. The USEPA is interested in a glossary, dictionary or lexical system rather than a structured set of terms. Having created this system, EPA realized that it didn’t have a good set of terms for cataloging, categorizing or sorting. There were three main terminology sources –a threelevel hierarchy used by the EPA Libraries for cataloging, a set of terms to catalog web pages and another to categorize datasets. Unfortunately, none of these resources had definitions. USEPA has embarked on a project with the states to create data standards in order to exchange information across geography and platforms. The state group has agreed to develop a standard scheme for the EPA for different materials and across states. In September EPA and the states will go through a political process to finalize the work. By adding this multi-level system, the TRS will need to become a more structured system. The Substance Registry System has over 73,000 chemicals that are regulated or tracked with CAS numbers, weights, structures and appropriate legislation. Other registries include Business Objects and Facilities. EPA is in the process of building a web service to submit a CAS number to EPA resources. EPA is still in the process of putting these services on the web because of technological issues. UN Initiatives Food and Agriculture Organization – Agrovoc Thesaurus (Margherita Sini, FAO, Rome) FAO has several terminology resources including Agrovoc which covers agriculture including forestry, fishers, nutrition, etc. in nine languages; FAOTERM which covers agriculture, biology, forestry, fisheries, economics, etc. in six languages; and AGRIS/CARIS, which is a categorization scheme of 130 categories at two levels in five languages. The latter has been mapped to Agrovoc. There are currently three ontology projects being developed to give more meaning and semantics to their web-based systems. The Fishery Ontology (with CNR), Food Safety, and, FNA (a bibliographic metadata ontology). The goal is to make ASFA, OneFish, Agrovoc and other terminologies interchangeable in order to have a global system and to allow searching across the four systems. A web site is available to demonstrate a search by synonyms, query expansion and term disambiguation. Multilingual support is included. A portal will be launched online soon to give semantic probabilities to the user. The Agriculture Ontology System (AOS) Project is aimed at creating a federation of information providers in the agricultural community with defined responsibilities and roles to develop a common methodology and interoperable KOS standards. FAO has collaborated on several terminology-related projects including a paper in the Journal of Digital Information (JODI), the SEMKOS project as part of the EU 6th Framework, and the organization of seminars and workshops in this area. IAEA - The INIS/ETDE Joint Thesaurus at the International Nuclear Information System (Yves Turgeon, IAEA, Vienna) INIS was created in 1969 as a cooperative system with 100 member states and 19 international organizations interested in the peaceful uses of nuclear energy. Its main product is a bibliographic database. The environment is an underlying theme in many of the subject areas; perhaps 14-20% of the material is related to the environment. Both the system’s categories and INIS/ETDE Joint Thesauurs are maintained in collaboration with the Energy Technology Data Exchange. Standard thesaurus relationships are used. INIS Dictionaries are available in English, French, German, Russian, and Chinese. Arabic is under development. Current developments include computer-assisted indexing and synonyms that work in the background during searches. The INIS Taxonomy is being revised based on a hierarchical category scheme because of the requirements for computer-assisted indexing. The first version is expected at the end of the year. UNEP - ENVOC Thesaurus (Gerard Cunningham, UNEP, Nairobi) The UNEP Thesaurus dates back to a glossary created in 1977. The Infoterra database was developed and then renamed to ENVOC in 1997. The target audiences are librarians, documentalists, database developers, and environmental information practitioners, primarily in developing countries. It is now available in 23 languages. ENVOC is a three-tiered hierarchy with the base terms grouped under a main category and subcategories. Content coverage is very general with highly aggregated terms. UNEP is planning to revise ENVOC this year by eliminating the numerical codes, adding definitions and solving the problems created by a categorized list of terms. UNEP is looking toward a global thesaurus but with a series of microthesauri. One would be a microthesaurus of macro, highly generalized terms. Others would be thesauri of micro terms for the specialist in particular areas. (The latter may be important to the use of terminology by the general public in developed countries where a regular thesaurus may be too complicated.) Good language coverage is required because it tends to break down the barriers to information sharing. UNESCO Thesaurus (Liane Barsony, UNESCO, Paris) The UNESCO Thesaurus is very multidisciplinary. It has seven large domains or microthesauri. Geographic groups are included in multiple groupings and this structure has been adopted by the World Tourism Organization Thesaurus. The alphabetical index is most often used by both indexers and end users. Each top term corresponds to a UNESCO program, service or section in the administrative structure and the work program. While it covers more than environmental subjects, there are over 1000 terms in the environment including those related to culture and educational environments. The UNESCO Thesaurus is used by the UNESCO library to catalog its holding and in collective cataloging projects with field offices, by outside organizations such as the UK Records Office, and in several integrated library management systems. The Albert Intelligent Search Engine, developed by UNESCO, incorporates the thesaurus based searching of descriptors and their synonyms. It is used to profile users to define subject interests, and the UNESCO Knowledge Portal will incorporate the thesaurus. JIAMCATT (Ali Benhadid, UNOG, Geneva) This is a network of members and national institutions for the sharing of terminology resources. JIAMCATT is a simple structure with an annual session and a Secretariat who manages the web site. Each partner contributes what it can, giving the partners responsibility to exchange expertise in the area of computer assisted technology and exchange of vocabularies. Working groups exchange views and developments among themselves and with software developers. The JAIMCATT community is larger than just linguistic service. Through the web site (a public part and members-only) access is provided to a common terminology repository which consists of a one-stop multi-site search developed by the EC. Right now they are not maintaining it so the links are broken. A file server is provided which archives the partners’ terminologies. All resources are in a common place to search; though they are investigating more distributed methods for the future. Other resources include a white board for discussions, links to other linguistic resources, and a directory of experts. JIAMCATT is also a coordination system. It proposes improvements and looks for inconsistencies. Terminology resources for databases are exchanged and a mechanism is provided for relaying information to decision makers at an annual meeting of the UN where language areas meet to discuss conference language and publishing services. WHO – CEHA Inter-Water Thesaurus and other WHO Sources for Health and Environment Terminology (Mazen Malkawi, WHO/CEHA, Jordan ) The Regional Centre for Environmental Health Activities in Amman, Jordan has the mandate to support the national program. The Centre supports the Eastern Mediterranean Region. CEHANET is the regional environmental health information network which has been in place since 1988 with the mandate to improve access to reliable relevant environmental health information in the region. The approach is to develop information systems and tools, produce regional and national information, develop human resources and establish environmental health information centres. The Water Thesaurus is not well known even by people in the organization. It is based on the IRC Inter-water Thesaurus, with water supply and sanitation terms added from the Netherlands. Materials are available in both Arabic and English. Regional and national databases are published as well as documents. The thesaurus is primarily a printed version because the UNESCO ISI package is being used. A mini-ISI electronic version is available. There are several subject categories including water supply, water quality, water use, etc. Equivalence, hierarchical and related relationships are included, along with scope notes. Thesaurus maintenance is of utmost importance but it is being done manually. There are focal points in the countries who use a form for proposing changes. Other support tools include a Subject Analysis Handbook, a procedure manual including ISO standards, and training courses. Approximately 100 trainers and 300 users in 23 countries have been trained. CEHA is looking for support and collaborators to finalize its environmental health specialized dictionary which will be in Arabic. A global classification scheme for the environment that is similar to MeSH would be very helpful. USE CASES AND USER NEEDS Towards a Web Service-Based Infrastructure: The Use of Environmental Thesauri within European Commission Initiatives (Paul Smits, EU/JRC, Italy) The DG Joint Research Centre (JRC) is part of the EC with the mission of providing customer-driven scientific and technical support for the conception, development, implementation and monitoring of EU policies. JRC functions as a reference centre for the EU. It is close to the policy making process with seven institutes in five countries. The Infrastructure for Spatial Information Europe (INSPIRE) has the goal of making harmonized spatial data available for policy formulation and integrating spatial data services based on a distributed network of databases linked by common standards and protocols. INSPIRE complements other key information directives including those on habitats, noise, and information policies such as copyright and data base protection. INSPIRE is about interoperability – how users access local, national or European level data resources. INSPIRE is in the process of being proposed to the EC with possible implementation by 2008. The pre-implementation phase is now underway. INSPIRE participates in a number of standards activities including the OpenGIS Consortium. However, standards alone are not enough; guidelines are required for unambiguous implementation of a standard. JRC is taking the lead in this area Two use cases were presented. The first case involves registering a resource. A current tool allows for selection of keywords from a thesaurus but the terms are from NASA. Another possibility is the UDDI Registry process where you can select from a controlled list to describe your web service. The use of UDDI registries should be exploited by the environmental community. Use Case 2 involves the INSPIRE EU portal, which is based on the metadata catalog. They would like to add thesauri to help guide the search. Thesauri, gazetteers, and other KOSs are central to JRC’s architecture. They want to ensure that efforts are not duplicated. Perhaps the ecoinformatics group could agree on a multilingual topic map for the environment with no copyright restrictions. The Ecoinformatic Initiative may want to contact the UDDI Consortium to make the topic maps available for use within the UDDI web services registry. The group should also agree on a protocol for web services, possibly based on SOAP, WSDL and UDDI. Other possible activities include linking up with software and services providers such as the UN GeoNetwork Initiative, encouraging producers of metadata tools to use thesauri rather than free text input, and feeding requirements, including multilingual considerations, into standards bodies in order to support the development of tools in standard ways. These actions would build an infrastructure that could be used by everyone. Discussion of Multilingual Considerations (Led by Gerard Cunningham, UNEP) Based on a GEMET example, with definitions sources and multiple languages, the group discussed the importance of multilinguality. One concept was a multilingual terminology reservoir which would be a glossary. A web service could be built around a public site that would provide access to such a glossary. These could be simple flat files with base terms and definitions in English. The InfoTerra partners could build a very simple system for distributing flat files of whatever terminology is available and then engage national partners (particularly governments and specialty institutes) to return their language equivalents of the terms with definitions. Translating and vetting authorities would need to be established to ensure quality. Multilingual resources such as GEMET can help to meet the needs of developing countries because they want to see their language against other UN languages. Environmental conventions tend to reflect a global environmental terminology, which is also important. It is important to put the impact of cultural differences on terminology as an open discussion point. A glossary could be viewed as a distinct product that would have multiple definitions to help identify specific cases of these cultural issues. A two-track approach may be needed. There may be a larger thesaurus effort along with a less structured effort that would provide multilingual resources for those languages that don’t have the wherewithal to do a large thesaurus. If a more granular, structured approach is necessary, there are a number of research issues. The Unified Medical Language System at the US National Library of Medicine produced a semantic network to which distinct KOSs could be mapped, but this is an extremely expensive project. The Use of GEMET for the Swiss Environmental Catalogue (Jean-Philippe Richard, GRID Geneva) Envirocat is a partnership between SAFL, the Swiss agency for environment, forests and landscape, and UNEP. There are two centralized metadata catalogues, Swiss and Alpine. GEMET has been chosen as the thesaurus because they wanted to save the amount of work invested in metadata collection and facilitate the importation of metadata entries which had already been applied. The availability of terms in German, French, Italian and English was also significant. Only a subset of GEMET is used in order to improve system performance. The size of the C-CDS terminology database is only 24,000 terms because of the restrictions to the four languages. A simplified database model has been employed which uses the BTs and NTs to create a relationship table. The data model retains the hierarchies and the attribution to EEA themes contained in GEMET. When indexing, GEMET terms are mandatory and the free text descriptors have been eliminated. The indexer can choose a single language or multiple languages. GEMET is also used in searching, including by topic. Several needs based on these uses were identified: microthesauri to solve the problem of GEMET’s granularity, adding or deleting terms to support local user needs, and implementing a tool to analyze how people are searching and how authors are using the thesaurus. It would also be beneficial to have a smaller thematic hierarchy. GEMET themes might be linked to theme lists from the Conventions or those developed for local uses. Could a thesaurus core be developed and maintained through an international group to which extensions for local needs could be added? Is it problematic to agree on a high level framework or to agree at the granular level – where can we find a middle ground? Do we have to agree or do we just have to show all the possibilities? Applying GEMET – UNEP’s Perspective (Sean Khan, UNEP) UNEP’s thesaurus applications include indexing and querying web-enabled database systems and quick profiling of expertise and institutions. Multilingual searches have offset translation costs. It is a way to get “a good enough” result, providing a good filter for what should really be translated by humans. Challenges include the multiple branch problem. For example, “biodiversity” gives you two examples, one under science and the other under biosphere. This is an issue when querying or displaying. A typical user will pick a single term. These broad top terms often cause problems because users pick the wrong tree. UNEP. Net is a framework consisting of two focused and distinct utilities: a discovery mechanism and support tools that allow users to use and share this information. It includes a simple user interface with a cartographic view, user defined themes, and filters by data type and source. Different schema are allowed; UNEP.Net will have to tackle the use of different thesauri particularly across different organizations. Typical users are secretaries who may not know the context of what they are trying to catalog. Several tools have been developed. One tool provides a list of all the related terms, synonyms, BTs and NTs from which the user can select the term to be cataloged. This has improved the indexing process. Users can also switch languages on the fly. Discussion of Needs of Different Client Groups The user needs are outlined in a report by the Austrian Federal Ministry of the Environment [1996]. The concept of a megathesaurus was discussed. Is it possible to devise a numbering or mapping scheme that would allow terms in one thesaurus to be mapped to those in another? This would allow searching of a resource that has been indexed using one thesaurus by using terms in another thesaurus. While this concept works on a small set, the question is whether it will scale. A high level ontology and some level of agreed upon facets might be an alternative to complete mappings. The possibility of linking specific microthesauri back to appropriate places in the trees of a more general thesaurus was also discussed. If the microthesaurus is not present then the user would have to do a follow on search using the term from the more general thesaurus. The question arose as to how the user or the system would know to use the fall back term. The definition of interoperability was discussed. Do we mean interoperability between the KOS contents or between databases? RELEVANT TECHNOLOGIES FOR INTEROPERABLE TERMINOLOGIES Demonstration of an Ecoinformatics Web Service (Tim Lynch) The Agriculture Network Information Center (AgNIC) is a collaborative venture among 40 institutions in the US led by the National Agricultural Library of the USDA with the goal of providing a single point of access to agricultural information. The site has a number of features including a search feature which is where the use of web services is currently focused. Instead of incorporating the thesaurus directly into the web site, they encouraged NAL to do it as a web service which provides more flexibility and makes the thesaurus more generally useful. The Ag Thesaurus is queried by the web service from the search box. The information in the database is categorized using terms from the thesaurus; the user does not have to know the preferred terms because the web service call to the thesaurus will pull them all together. Mr. Lynch demonstrated the linkage that is possible between two terminologies using web services. He searched a German term in GEMET, identified the English translation, and then searched the AgThesaurus and the AgNIC resources. In order to do this he cloned the software from the AgNIC web service and loaded a portion of the GEMET database in English, Spanish and German. It required that GEMET be set up as a web service and then about 10 lines of code were added. Web service protocols for thesauri are needed. Work has been done in this area by the Alexandria Digital Library. However, problems were found when requesting a hierarchy of terms with some environments such as Coldfusion. The AgNIC protocol was adapted from the ADL to work with a variety of web service technologies. The protocol supports requests for BTs, NTs, a number of levels, and for a known term or term-based pattern matching. The relevant issue is how to structure the query. ISO 5963, the specification for searching a thesaurus, is an approach that could be used. Issues of branding have not been addressed. AgNIC will be mirroring the Ag Thesaurus. It is asking users of their services to register with them, and they credit services that they are consuming on the AgNIC site. Questions of performance and reliability of services or resources that you do not own also exist. There are several other web services used on the AgNIC site including a common calendar feature, the featured site feed from an LDAP-like database, and the browsing by topic. Suggestions on Sharing Environmental Terminologies (Stefan Jensen, EEA) Mr. Jensen proposed setting up GEMET maintenance as a WIKI to address practical problems of maintaining GEMET in a centralized fashion. In looking for technologies to support sharing and maintenance of a resource, Mr. Jensen came across WIKI which is used in open resources like Wikipedia and Wik-Dictionary. Zope-Openwiki would allow such an open structure to be built. It would make GEMET an open web-source for others to take from and add to. This is especially important for a multilingual resource when it is no longer feasible to do the expansion through specific physical groups. GEMET could be used as a core terminology and users could add to it with some moderation to assure quality. Copyright would not be retained but a GNU public license could be used. A disclaimer would need to be considered. Issues include validation and authoritative sourcing. A persistent version of GEMET may be needed for some applications. It was suggested that GEMET be implemented as a WIKI on an experimental basis, as long as the original GEMET is maintained in a stable way and the differences between the two versions are made clear. Mondeca offered to host such a WIKI environment to try out this concept since Mondeca has already converted GEMET. Thesaurus Software Tools - THESmain, THESshow and SuperThes (Hermann Stallbaumer, TBHS, Austria) SuperThes software is the latest version of THESmain which dates back to 1995. This software product is used by ENVOC, UDK-Thesaurus, GEMET, AWG, etc. Several problems have surfaced. GEMET now has 19 languages and there has always been a problem incorporating new languages, including issues related to character sets. Thesaurus constructors demand flexible thesaurus structures, the freedom to create thesauri from scratch or to import them from various places, convenient data exchange with standard office applications using state-of-the-art technologies like drag and drop, report generation, and bulk extraction using standard file formats such as XML. SuperThes is being implemented as a web service. ODBC and JDBC connections are available which are open source. SuperThes currently uses an XML DTD. XML schema are required for web services, but there are automatic tools that can be used to convert DTDs to schema. The relationship types have been expanded. Integrating Ontology and Thesaurus in a Semantic Web Framework Use Case: Multimedia Dictionary of Sustainable Development (Bernard Vatant, Mondeca) Mondeca is a small French company that has been developing KM solutions based on concept-like semantic nets and topic maps. Mondeca tries to stay compatible with the Semantic Web. Topic maps are not considered part of the SW framework according to the W3C, but this is changing. In this environment some “things” you are interested in are reflected in terminologies (subjects) while others such as people are generally managed as part of taxonomies. The third component is the information resource (document and data). Based on these objects, there are several large questions. How do you make these three domains come together? How do we agree we are speaking about the same thing either within the same KOS or across KOSs? How do we access distributed information about something? Even if we agree about the “thing”, how do we agree on classes and properties of that thing in an ontology? How formal must the agreement be to disagree – is the agreement for humans or computer? How do we relate “things” to each other? How does metadata definition relate to the individual instances that are cataloged? The Semantic Web tries to address these issue using web technology. Everything is represented as a URI and everything is a resource (RDF). The URIs identify resources. OWL provides a standard way to declare and commit to an ontology based on RDF-XML and RDF semantics. RDF allows you to link resources together via semantic predicates. Topic Maps bring features that are not native to RDF+OWL. These include multilingual management using name and scope mechanisms and subject indicators which clarify the subject a URI actually identifies. However there is still debate about how a subject indicator defines identify. The Topic-Role Association structure of topic maps allows multiple relationships to be identified and the natural expression of complex relationships that occur in knowledge bases to be described. However, topic maps do not have built in formal semantics which RDF and OWL have. A thesaurus is a sort of ontology but with weak formal definitions of the relationships which are really needed to do a well-formed taxonomy. Formal semantics are needed if you want to have computer action against the ontology. A question that is often asked is how legacy thesauri can be ported to the semantic web. How can implicit or explicit thesaurus semantics be expressed? What languages are best to express this? There is ongoing work on these issues within SWAD-Europe and the Semantic Web Best Practices and Deployment Working Group of the W3C. Thesauri should be integrated into the SW framework; it should not be thrown out or lost in the RDF soup. In an Ontology Driven Thesaurus, the current thesaurus can be presented and constraints can be added (this works especially if you have thesaurus terms classified). A prototype project by Planete-Ecologie Directory of Web Resources uses GEMET. Mondeca has two years of experience in dealing with thesauri in this manner and they now have a full scale project called Dictionnarie Multimedia du Developpement Durable (in French) to convert multiple thesauri. Semantic Web Advanced Development for Europe Project: SWAD-E (Alistair Miles, Rutherford Appleton Lab., UK) The Web is currently information for humans, while the Semantic Web is the development of a Web that has data so that computers can understand meaning. The goal is to improve searching and to have the Web do more of what humans must do now. The enabling technologies are RDF and OWL. SWAD-E is an EU project set up to support the W3C Semantic Web activities through R&D, demonstrations, application development and the identification of guidelines. SWAD-E is interested in thesauri because they represent large bodies of well-engineered knowledge that the semantic web can bootstrap from. SWAD-E is designing RDF schemas to express thesaurus data, developing guidelines for migration and use of thesauri, developing supporting technologies, and demonstrating the value of using thesauri in this context. RDF allows “chunks” of a thesaurus to be made available on the Web. A link can be created while changing and evolving the connected data structure. The move to XML is a step in the right direction but XML is not equal to interoperability; there are no two thesauri on the web in the same XML format. Because of this SWAD-E has developed the SKOS-Core 1.0, an RDF schema for thesauri. There are several challenges related to the SKOS schema. There is a need to preserve unique features of individual thesauri while supporting interoperability. The second challenge is interoperating while migrating all the different types of KOSs. Therefore, the underlying SKOS meta-model is concept oriented rather than term oriented. URIs are assigned to concepts, and a hierarchy of extensible semantic relations is established. Semantic mappings are created to identify the relationship between the concept in one schema and the concept in another. Using GEMET as an example, a few extensions would be needed to handle the nonstandard features such as groups, super-groups and themes. These would be additional concepts that would have a certain type of relationship. SWAD-E is planning to use a combination of SKOS Core and OWL to represent this hybrid structure which is half way between a thesaurus and an ontology. Later, the relationships can be extended. An SKOS Mapping supports the interchange of thesauri. Using the SKOS APIs an interface could be built to terminology web services for searching, querying, and browsing. A web service API could be built for the environmental terminology community, so that we can talk to one another. A reference implementation of such an API is under development. Issues such as versioning and evolution of the thesaurus need to be discussed. SWAD-E would be interested in working with the environmental terminology community to transform thesauri for the Semantic Web. The SKOS Core 1.0 Guidelines for Migration are under development. An XML exchange format is needed. A web service that gets a description of an individual terminology resource is also important. NEXT STEPS, PROPOSALS AND RESEARCH AGENDA Geosciences Proposal (Jan Jellema) A proposal is being written for the EU to develop tools for geo-terminology (GEOTERM). The proposal will define the organization needed to support such an effort and the guidelines for the content. The project would provide an Internet presence for terminology in the geosciences and relate this terminology to other thesauri. GEMET might be the top level and the geo terms could be linked at a lower level. Through an XML schema for each language, this framework would create and maintain the thesaurus in a decentralized manner and allow for more automated updating. National and special extensions would be needed in such a framework since geosciences terminology is by its nature geographically based. A public interface could be developed to provide popular terms while continuing to serve the expert’s needs. Groups of international and national experts would oversee GEOTERM’s development. The thesaurus would integrate more easily with other applications. An HTML page could be made for each term in the terminology. General applications such as adding keywords into a Word document would be more feasible. Web services for search, translation and XML import and export would be needed. Discussion of Future Cooperation on Environmental Terminology Based on the presentations by the various participants, it became clear that there are a variety of KOSs that are applicable to the environment. Something is needed that will bring all these resources together. One could view this as “web service choreography”. The group decided that it is necessary to start with a vision, to develop a list of components and a roadmap to achieve them, and then to set up a testbed for performance purposes. The first issue is that of one-to-one web services. The group also acknowledged the need for visibility in other communities and a need for funding. A group name, ECOTERM, was suggested. Several pilot projects were discussed. The possibility of a follow-up meeting next year was of interest to those in attendance. SUMMARIZATION OF RECOMMENDATIONS AND NEXT STEPS The following is a draft work plan for future cooperation on Environmental Knowledge Organization Systems. The action items are grouped by administrative, technical, marketing and communications and other. Note: KOSs include thesauri, classification schemes, authority files or any other formal terminology system. [put a more formal definition from the CLIR report] ADMINISTRATIVE Agreed the name of this group and project is Ecoterm •Need statement of common vision (Gerry to draft) •UNEP to finalize and distribute meeting report (Gerry Cunningham) •Plan an annual Ecoterm meeting possibly in conjunction with a UNEP.Net meeting (Gerry Cunningham) TECHNICAL •Need for use cases was identified and are requested from anyone who will submit them (they should include specific scenarios, needed KOSs and user audiences, if possible) –E.g., KOSs for the European Spatial Data Infrastructure; use of existing KOSs in Egovernment-wide taxonomy development; retrieving definitions from multiple KOSs) •Develop example implementations of a thesaurus web service •Define high level services that are needed •SKOS reference implementation (Tim Lynch) •GEneral Multilingual Environmental Thesaurus (GEMET) as a web service (Sean Khan and Stefan Jensen) •EEA will liase with SWAD-E project to get the full GEMET available in their SKOS format •EEA will follow up with Tim Lynch (USDA, Cornell) during 2nd week of May to jointly work on the prototype web-service done around GEMET in ZOPE. This can either include the full coverage of GEMET in the USDA web-service or the finishing of the work through EEA consultants. •Consider exchange between web-services based on the USDA prototype through SOAP or XML/RPC or both •EEA will elaborate on the hosting of this web-service on one of the EIONET servers and according to that decision provide the domain part of the URI for all GEMET terms •Review of options related to technologies •“Standard” formats for KOSs on the web •Identification of APIs and requirements •Investigate the feasibility of registering KOSs in a UDDI (Universal Description, Discovery and Integration) Registry (Thomas Bandholtz) •Consider the feasibility of publishing RDF encodings of KOSs on the Web •Review the taxonomy and definitions of relationship types with SKOS (Gail Hodge, Bernard Vatant, Margharita Sini and Alistair Miles) •Determine what information should come back for a resource that would help the user to evaluate multiple resources -- e.g., branding, quality, context/scope parameter, etc. •Develop a list of standards and other resources related to KOSs, (e.g., the Alexandria Digital Library Guidelines for Web Services) (Bruce Bargmeyer, Paul Smits and Gail Hodge) MARKETING AND COMMUNICATIONS •Find a mechanism for sharing information within the group •Update the terminology page on the Ecoinformatics site including the rename of the project to Ecoterm •Consider use of the bulletin board on the Ecoinformatics page for communications within this group (Stefan Jensen) •Put up the presentations from the meeting (Gerry Cunningham) •Promote our KOSs with groups involved in new technologies such as the semantic web •Form liaison with W3C Semantic Web Activity - Best Practices and Deployment Working Group (Bruce Bargmeyer, Bernard Vatant and Alistair Miles) • Promote our terminologies with the Earth Observation Summit (GEO) (Paul Smits) Send existing questionnaire to stakeholder groups who weren't at the meeting and ask those who were invited but could not come to submit questionnaire responses (Gerry Cunningham and Paolo Plini) Distribute Draft Content Standard for KOS Description (Hill & Hodge) to the Ecoterm group for comments and modify as needed (Gail Hodge) Consider a more extensive questionnaire based on the revised Content Standard (Gail Hodge) Enter the questionnaire information into a database (terminology registry) (Sean Khan and Gail Hodge) Put URLs and descriptions for those resources that are web accessible on the Ecoterm web page (Gail Hodge and Stefan Jensen) Extend ENVOC (UNEP's Environmental Vocabulary) and GEMET to additional languages (Gerry Cunningham and Stefan Jensen) Set up an e-mail reflector with an archive including everyone from the meeting (Gerry Cunningham or Stefan Jensen) (The reflector should put the name of the group in the beginning of the subject line; followed by very short but understandable thread name and subject) Increase the visibility of environmental KOSs by investigating ways to make individual terms accessible directly via a Google-like search OTHER •Geosciences Proposal to the EU on sharing -- sharing of tools •Explore the possibility of making GEMET publicly available through WIKI technologies. This may provide a maintenance and look-up environment. (Stefan Jensen and Bernard Vatant) REFERENCES Federal Ministry for the Environment (Austria) (1996). User Experiences with Environment Thesauri in CDS. First International Workshop: Catalogue of Data Sources (CDS) and Thesaurus. Publication Series Umweltdatenkatalog, Volume 8. Vienna, Austria, April 1996. Hill, Linda & Gail Hodge. Taxonomy of Knowledge Organization Sources/Systems. NKOS. July 2000. (http://nkos.slis.kent.edu/KOS_taxonomy.htm ) Hodge, Gail. “Systems of Knowledge Organization for Digital Libraries: Beyond Traditional Authority Files” CLIR Pub91. April 2000. (www.clir.org/pubs/abstract/pub91abst.html ) Appendix A Environmental Thesaurus/Terminology Workshop United Nations Environment Programme (UNEP) International Environment House (UNEP Regional Office for Europe) 9 – 15, chemin des Anémones CH-1219 Châtelaine, Geneva, Switzerland 14-15 April 2004 Agenda Purpose: Bring together the major providers of environmental terminologies to discuss the status of their terminologies, how they are applying new technologies, and how these resources can be “integrated” using new technologies. Objectives: Bring together major providers of environmental thesauri and provide a forum for discussion about the current use and future of their resources Present the concept of web services and demonstrate a prototype of several simple services Facilitate discussion among the participants to determine whether this approach can be useful and if other groups are interested in implementing these services, developing new ones, etc. Collect information from participants that can lead to an enhanced terminology web page through the collection of key information about each terminology Present key components of terminology management including the terminology types and the concept of different behaviors that influence how and what services are provided Identify a research agenda and possible funding opportunities to move the process forward Create an outcome relevant to one or more of the topical focus areas identified as being of interest to the EPA and EEA – children’s health, biodiversity, chemicals, etc. so that the results can be presented at the Ministerial Meeting in May. Wednesday, 14 April 2004 Workshop Introduction Session 1 – Chair: Gerard Cunningham, UNEP 0900h Welcome and Opening Remarks (Frits Schlingemann, Director, UNEP/ROE) 0915h Participant Introductions 0945h Ecoinformatics partnership and objectives of the workshop (Bruce Bargmeyer/Larry Fitzwater, USEPA & Stefan Jensen, EEA) 1000h Types and varieties of terminology resources (Gail Hodge, USGS/NBII) 1010h Introduction of web services concept (Tim Lynch, USDA) 1030h Coffee break Overview of Existing Systems and Use Scenarios 1100h Overview of existing environmental thesaurus/terminology initiatives (Participants to prepare one-page overviews using a questionnaire) 1100h Environmental Applications Reference Thesaurus – EARTh (Paolo Plini, CNR, Italy) UDK Thesaurus and UMTHES (Wolf-Dieter Batschi, UBA Germany) TED - Thesaurus on Emergencies and Disasters (Rudolf Legat, UBA Austria) 1110h 1120h 1130h 1140h 1150h National Biological Information Infrastructure – NBII (Lisa Zolly & Vivian Hutchison, USGS) CAB Thesaurus (James Brooks, CABI) Integrated Taxonomic Information System – ITIS (Janet Gomon, Smithsonian Institution, USA) 1200h SNS – The Semantic Network Service (Maria Ruether, UBA-Germany & Thomas Bandholtz, Consultant) 1210h Environmental Multilingual Thesaurus of the Spanish Environment Ministry (Carmen Casal Fornos & Arantza López de Sosoaga, MMA) 1220h Lunch Continuation of the presentations Session 2 – Chair: Gail Hodge, USGS 1320h Terminology Co-ordination and Harmonisation in the EU in the light of the IATE Project (Dieter Rummel, EU ) 1330h Multilingual Thesaurus for the GeoSciences (Jan Jellema, TNO-NITG, Netherlands) 1340h GEMET and EEA Glossary (Stefan Jensen, EEA) 1350h USEPA TRS (Larry Fitzwater, USEPA) 1400h UN initiatives: FAO – Agrovoc thesaurus (Margherita Sini, FAO, Rome) IAEA - The INIS/ETDE Joint Thesaurus at the International Nuclear Information System (Yves Turgeon, IAEA, Vienna) UNEP - ENVOC Thesaurus (Gerard Cunningham, UNEP, Nairobi) UNESCO Thesaurus (Liane Barsony, UNESCO, Paris) UNITED NATIONS - JIAMCATT (Ali Benhadid, UNOG, Geneva) WHO – CEHA Inter-Water Thesaurus and other WHO Sources for Health and Environment Terminology (Mazen Malkawi, WHO/CEHA, Jordan ) 1500h Coffee break Use Cases and User Needs 1530h Towards a web service-based infrastructure: the use of environmental thesauri within European Commission initiatives (Paul Smits, EU/JRC, Italy) 1600h Multilingual considerations – discussion 1615h The use of GEMET for Swiss environmental catalogue (Jean-Philippe Richard, GRID-Geneva) 1630h Applying GEMET – UNEP’s perspective (Sean Khan, UNEP) 1645h Needs of different client groups - discussion 1800h 1900h Adjourn Reception (La Terrassa restaurant) Thursday 15 April Session 3 – Chair: Stefan Jensen, EEA Relevant Technologies for Interoperable Terminologies 0900h Demonstration of Ecoinformatics Prototype (Tim Lynch, USDA) 0945h Thesaurus Software Tools - THESmain, THESshow and SuperThes (Hermann Stallbaumer, TBHS, Austria) 1015h Coffee break Presentation of Relevant Technologies contd. 1045h Integrating Ontology and Thesaurus in a Semantic Web Framework Use case : Multimedia Dictionary of Sustainable Development (Bernard Vatant, Mondeca) 1115h Semantic Web Advanced Development for Europe project: SWAD-E (Alistair Miles, Rutherford Appleton Lab., UK) Technology options 1145h Some suggestions on sharing environmental terminologies (Stefan Jensen, EEA) 1215h Lunch Next Steps, Proposals and Research Agenda contd. Session 4 – Chair: Bruce Bargmeyer & Larry Fitzwater, USEPA 1315h Geosciences Proposal (Jan Jellema, TNO-NITG, Netherlands) 1345h Future cooperation on environmental terminology: (discussion session aimed at reaching agreement on respective organizational roles on future terminology collaboration) 1500h Coffee break 1530h Future cooperation on environmental terminology (continuation) Closing session 1630h 1730h Summarization of recommendations and next steps Concluding remarks Closure of the workshop (UNEP representative) Adjourn Appendix B List of Participants Appendix C Terminology Survey Appendix D Survey Responses Appendix E Draft Agreement