Re: [unrev-II] Towards a summary of documents (including Conceptual Graphs based on existential graphs of Charles Peirce)

From: John J. Deneen (
Date: Mon Nov 06 2000 - 13:27:39 PST

  • Next message: John J. Deneen: "Re: [unrev-II] Towards a summary of documents (including Conceptual Graphs based on existential graphs of Charles Peirce)"

    Here's another good reference:

    Building and Searching an XML-based Corporate Memory
    (Auguste Rabarijaona, Rose Dieng, Olivier Corby, INRIA, France)

    Abstract: This paper emphasizes the interest of XML meta-language for corporate
    KM. Taking into account the advantages of the Web and ontologies for KM, we
    present OSIRIX, a tool enabling enterprise-ontology-guided search in XML
    documents that may consitute a part of corporate memory.

    "Advantages of XML for a corporate memory

       * Many Views on the Same
         XML enables to manage information and knowledge in a unique structured way
         and enables several different processings. Knowledge servers retrieve
         information while clients are in charge of presenting it to users through
         adapted interfaces. It is then possible to take users and context into
         account and to present different views of the same data: it may be possible
         to generate graphic views, table of contents or to show the data
         themselves. Furthermore, data are loaded into the client (the browser) and
         can be processed locally: e.g. XML data can be processed by Java applets
         [5]. Hence, XML may represent for data what Java represents for programs:
         transparent portability through machines and operating systems.
       * Documents Built from Heterogenous
         XML enables to manage structured documents and structured data in a uniform
         way. The XML format has been designed to enable document description as
         well as arbitrary data description. It is hence possible to mix data and
         documents in order to build virtual documents issued from several sources.
         Data may come from a technical data base while text may come from a
         document management environment. Furthermore, it is possible to annotate
         documents with modeled knowledge, so­called ontologies.
       * Standard for Information
         In order to facilitate communication and information exchange, a community
         (i.e. a department, a company, a group of companies of the same domain, a
         company and its related providers and clients, etc) may define a standard
         domain­oriented or application­oriented vocabulary by means of a DTD
         (Document Type Definition). A DTD is a syntactic specification being used
         as model for XML documents. A document is considered as valid if it
         respects the DTD with which it is associated. Documents or data can then be
         expressed with the defined XML markups and then be exchanged using these
         markups [5].
       * Document
         XML has a companion formalism called XSL (Extensible Stylesheet Language),
         to define document­oriented presentation format. XSL may present a document
         in HTML, PDF, etc. It may also generate a generic format, that may be
         postprocessed to generate a standard output format. XSL also enables
         elementary document processing such as sorting, generating table of
         contents, tables, reorganizing the document structure. Using XSL, it is
         hence possible to define several output formats for the same document
         structure: XSL is a document transformation and formatting language. It is
         possible to write once and to publish many times, from the same source, to
         different media: digital and paper­based ones. This is very interesting for
         corporate memory management.
       * Hypertext.
         XML will also offer tools to build powerful hypertext documents by means of
         XLL (XML Linking Language), and XPointer, the language that enables
         navigation in documents according to their structure. XLL will implement
         the major hypertext functionality that can be found in dedicated tools:
         links between more than two documents, external links, links with
         semantics, etc. With external links, documents can be annotated from the
         outside, without modifying the source.
       * Information
         XML facilitates information search because documents are structured and,
         hence, can be considered as a database. It is possible to rely on
         standardized markups to search information in a structured way. Moreover,
         the database community is currently integrating XML with database
         technology and search languages (cf. XML­QL [10]).
       * Annotation with
         In the spirit of RDF [16], Shoe [18] and OntoBroker [13], it is possible to
         annotate a document with metadata which describe in an abstract and
         synthetic way what the document is about, according to predefined standard
       * Semantic
         It is also possible to markup parts of the text document with tags the
         semantics of which is to establish a relationship with ontologies and
         knowledge models. In some cases, the markup can be interpreted as a
         hypertext link to ontologies. It can also be exploited by markup­driven
         search engines.
       * XML and Memory
         XML as a structured document open standard may be a good candidate to
         facilitate migration to new systems or software through long time period.
         With XML, documents exist by themselves, independently of the processing
         tools. There are no more X or Y proprietary format documents but just (XML)

    Thanks to all these features, XML is an interesting meta­language for building a
    document-based corporate memory aimed at being accessible to the Intranet of a

    5. Enterprise model ­ guided search in XML documents
    Taking into account the interesting features of XML, we developed techniques for
    enterprise model­guided search in XML documents.

    5.1 Information search guided by knowledge models
    5.1.1 Objectives
    Our main objective is to perform information search in documents on the Web,
    guided by knowledge models. The result of search should include only relevant
    answers i.e. Web documents which «correspond semantically» to the request.

    Instead of developing a specific extension of HTML to support a given ontology
    (as for example, Ontobroker or Shoe), we choose to handle XML documents,
    formatted with XML standard. The knowledge models that will guide the search
    will be CommonKADS expertise models, represented in standard CommonKADS
    Modelling Language (CML) 1 .

    The system OSIRIX (Ontology­guided Search for Information Retrieval In
    XML­documents) is integrated in the following global architecture:"

    (see Fig. 1 : Global architecture.)

    Jack Park wrote:

    > kMail is problematic. The paper cited is valuable in that it outlines the
    > issues involved in the construction of an organizational memory (a NIC).
    > Unfortunately, it does not appear to be possible to track kmail further
    > because there exists a KMail that is part of the KDE Linux desktop system.
    > That seems to hog all the kmail hits at google.
    > Conceptual Graphs were the creation of John F. Sowa, based, he says, on C.S.
    > Peirce's Existential Graphs. Sowa may not follow Peirce to the letter, but
    > his work is gaining a grand following, not to mention, an email list devoted
    > to CGs.
    > In view of the existence of major events related to the Semantic Web, it
    > would seem that this post is very timely and has much to offer the Unrev
    > group in relation to the OHS project. Here is an accumulation of URLs that
    > relate to this post:
    > a new organization
    > SW technologies workshop
    > semantic web portal community
    > t <<watch out, that belongs with the URL> semantic web -- a primer
    > narval
    > a narval child
    > collab. virtual workspace
    > OHS project
    > next year's conceptual structures
    > workshop
    > From: John J. Deneen <>
    > To: <>
    > Sent: Friday, November 03, 2000 12:17 PM
    > > I'm sure you'll find the following links about 1) kMail, 2) WebKB and 3)
    > > Conceptual Graphs based on the existential graphs of Charles Sanders
    > Peirce as
    > > enabling solutions for powerfully summarizing our Unrev2 mailing list!!!
    > >
    > > 1) kMail: E-mail, contexts and views
    > > In "Tying Knowledge to Action with kMail, " David G. Schwartz and Dov
    > Te'eni
    > > exploit the Internet and e-mail to disseminate knowledge for the purposes
    > of
    > > capturing Organization Memory (OM) and "collective IQ". They emphasis the
    > need
    > > to tie knowledge to action and present the kMail system, which integrates
    > e-mail
    > > with organizational memories to distribute knowledge over the Internet at
    > the
    > > right time and to the right person. kMail lets e-mail senders
    > contextualize
    > > their messages by creating revelant views of the organizational memory. It
    > also
    > > uses metaknowledge consisting of user profile information and shared
    > semantics
    > > information. They describe the context through shared metaknowledge and
    > link it
    > > to the organization memory each time knowledge is used to perform an
    > action.
    > > Domain-specific criteria help determine the revelant views in the
    > organization.
    > >
    > > IEEE Intelligent Systems, Vol. 15, No. 3, May/June 2000
    > >
    > >
    > > 2) WebKB: Ontology-guided knowledge retrieval with conceptual graphs for
    > > representing metadata on Web documents
    > >
    > >
    > >
    > >
    > > 3) Conceptual Graphs
    > > Conceptual graphs (CGs) are a system of logic based on the existential
    > graphs of
    > > Charles Sanders Peirce and the semantic networks of artificial
    > intelligence.
    > > They express meaning in a form that is logically precise, humanly
    > readable, and
    > > computationally tractable. With their direct mapping to language,
    > conceptual
    > > graphs serve as an intermediate language for translating computer-oriented
    > > formalisms to and from natural languages. With their graphic
    > representation,
    > > they serve as a readable,
    > > but formal design and specification language. CGs have been implemented in
    > a
    > > variety of projects for information retrieval, database design, expert
    > systems,
    > > and natural language processing.
    > >
    > >
    > >
    > > References
    > > Slides and demo of the WebKB set of tools
    > >
    > >
    > >
    > > Proceedings of Tenth Knowledge Acquisition for Knowledge-Based Systems
    > Workshop
    > >
    > >
    > > The Enterprise Ontology
    > > The Enterprise Ontology is a collection of terms and definitions relevant
    > to
    > > business enterprises. The ontology was developed in the Enterprise Project
    > by
    > > the Artificial Intelligence Applications Institute at the University of
    > > Edinburgh with its partners: IBM, Lloyd's Register, Logica UK Limited, and
    > > Unilever. The project was support by the UK's Department of Trade and
    > Industry
    > > under the Intelligent Systems Integration Programme(project no
    > IED4/1/8032)
    > >
    > > Conceptually, the Enterprise Ontology it is divided into a number of main
    > > sections -- these are summarised below.
    > >
    > >
    > > Other Publications, etc.
    > >
    > >
    > ============================================================================
    > This message is intended only for the use of the Addressee(s) and may
    > contain information that is PRIVILEGED and CONFIDENTIAL. If you are not
    > the intended recipient, dissemination of this communication is prohibited.
    > If you have received this communication in error, please erase all copies
    > of the message and its attachments and notify
    > immediately.
    > ============================================================================
    > Community email addresses:
    > Post message:
    > Subscribe:
    > Unsubscribe:
    > List owner:
    > Shortcut URL to this page:

    This archive was generated by hypermail 2b29 : Mon Nov 06 2000 - 13:37:53 PST