Virtual annotation fosters understandingBy Eric Smalley, Technology Research News
A set of software tools that allows people to record knowledge of a subject could make searching through documents more like questioning an expert than searching for a string of keywords.
Researchers at The Open University in England are applying techniques from the knowledge representation branch of the artificial intelligence field to the everyday task of working with electronic documents.
"Document-centered work is a fact of life," said Enrico Motta, director of the Knowledge Media Institute at the Open University. "Much of the knowledge work in an organization requires producing, retrieving and discussing documents. We are exploring ways to facilitate knowledge sharing and creation centered around documents," he said.
The key to the approach is it allows people to create a shared conceptual framework as a reference point for accessing the information in collections of documents.
"Our technology makes it possible to rapidly augment web sites with discussion spaces and conceptual models," Motta said.
These conceptual models are based on ontologies. "An ontology [is] a shared terminology which can be used to make statements about some domain," said Motta. "For instance, if I want to describe what goes on at the Knowledge Media Institute, my ontology would include generic concepts such as ‘academic’, ‘project’, ‘technology’, relations such as ‘x is involved in project y’, as well as concrete and abstract individuals such as ‘Enrico Motta’ and ‘Project ScholOnto’," he said.
Unlike most shared bodies of knowledge, these ontologies are published specifications, Motta said.
For example, a library of scientific papers and a conceptual model which describes the various theories, approaches, technologies described in the papers make it possible to find answers to questions such as "Has anybody built on the results produced by Bloggs & Bloggs," or "Have data been published in the literature which go against the hypothesis formulated by White & Black?" said Motta. "These questions cannot simply be answered by lexical or keyword-driven search methods. They reflect the interpretations people... impose on a library," he added.
The idea of using conceptual models to organize and access information has been around for more than a decade but has only led to a wide range of research projects in the last year, said Deborah L. McGuinness, associate director of the Knowledge Systems Laboratory at Stanford University. "That field is essentially taking off at the moment," she said.
For example, the Defense Advanced Research Projects Agency's (DARPA's) Rapid Knowledge Formation initiative aims to let non-technical people build knowledge bases. The initiative is supporting research projects at Cycorp, SRI International, Information Extraction & Transport, Inc., Northwestern University, Stanford University, University of West Florida, Massachusetts Institute of Technology, Pragati Synergetic Research, Inc., George Mason University and Teknowledge Corp.
It's a good idea to define terms, make them centrally available and organize them hierarchically, said McGuinness. This type of organization means that "you have more general terms like ‘animal’ above more specific terms like ‘cat.’ So a cat would inherent all the properties an animal has plus have some additional ones related to being a cat," she said.
The Open University system organizes the information contained in documents, but leaves the documents themselves untouched by annotating them virtually. This allows more than one conceptual model to organize a given set of documents. A conceptual model can also organize several sets of documents at once.
The system is composed of four tools. The Operational Conceptual Modeling Language is for building ontologies. WebOnto is an ontology browsing and editing tool. Lois is a form that allows users to query an ontology and Knote is a form that allows users to add information to an ontology.
The researchers have used the system to organize library information and to organize news into personalized news feeds based on a conceptual model of the users' activities and interests, said Motta. The researchers are working on a dozen projects aimed at further developing the software and deploying the technology in real-world applications, he said.
Motta's research colleagues were John Domingue, Simon Buckingham Shum, Zdenek Zdrahal and Paul Mulholland. Motta, Domingue and Shum published a paper about the research in the June 2, 2000 issue of the International Journal of Human-Computer Studies. The research was funded by the Engineering and Physical Sciences Research Council, the European Community, the National Health Service, the ADAPT programme, BP Amoco, Marconi Corp. and INNN Ltd., Motta said.
Funding: Government, Corporate
TRN Categories: Databases and Information Retrieval
Story Type: News
Related Elements: Technical paper, "Ontology-driven document enrichment: principles, tools and applications," International Journal of Human-Computer Studies, June 2, 2000
December 20/27, 2000
Tiny wires store more
Tinier transistors keep Moore's Law on track
3-D geometry adds twists to microfabrication
Virtual annotation fosters understanding
Silicon process produces pockets
Research News Roundup
Research Watch blog
View from the High Ground Q&A
How It Works
News | Blog | Books
Buy an ad link
Ad links: Clear History
Buy an ad link
© Copyright Technology Research News, LLC 2000-2006. All rights reserved.