The case for explicit knowledge in documents

Leslie Carr, Timothy Miles-Board, Arouna Woukeu, Gary Wills, Wendy Hall
2004 Proceedings of the 2004 ACM symposium on Document engineering - DocEng '04  
The Web is full of documents which must be interpreted by human readers and by software agents (search engines, recommender systems, clustering processes etc.). Although Web standards have addressed format obfuscation by using XML schemas and stylesheets to specify unambiguous structure and presentation semantics, interpretation is still hampered by the fundamental ambiguity of information in #PCDATA text. Even the most easily distinguishable kinds of knowledge such as article citations and
more » ... er nouns (referring to people, organisations, projects, products, technical concepts) have to be identified by fallible, post-hoc extraction processes. The WiCK project has investigated the writing process in a Semantic Web environment where knowledge services exist and actively assist the author. In this paper we discuss the need to make knowledge an explicit part of the document representation and the advantages and disadvantages of this step.
doi:10.1145/1030397.1030417 dblp:conf/doceng/CarrMWWH04 fatcat:hjrmja6q2nfp3ke3i3dieortqm