[Semantic Web]Implicit Semantics

读书笔记,  电脑与网络 Lee 发表于 2006/5/17 15:10:28 This type of semantics refers to the kind that is implicit from the patterns in data and that is not represented explicitly in any strict machine processable syntax. Examples of this sort of semantics are the kind implied in the following scenarios:   • Co-occurrence of documents or terms in the same cluster after a clustering process based on some similarity measure is completed. • A document linked to another document via a hyperlink, potentially associating semantic metadata describing the concepts that relate the two documents. • The sort of semantics implied by two documents belonging to categories that are siblings of each other in a concept hierarchy. • Automatic classification of a document to broadly indicate what a document is about with respect to a chosen taxonomy. Further, use the implied semantics to disambiguate (does the word “palm” in a document refer to a palm tree, the palm of your hand, or a palm-top computer?). • Bioinformatics applications that exploit patterns like sequence alignment, secondary and tertiary protein structure analysis, and so forth.   One may argue that although there is no strict syntactic and explicit representation, the knowledge about patterns in data may yet be machine processable. For instance, it is possible to get a numeric similarity judgment between documents in a corpus. Although this is possible, this is the only sort of processing possible. It is not possible to look at documents and automatically infer the presence of a named relationship between concepts in the documents. Even though the exploitation of implicit semantics draws upon well-known statistical techniques, the wording is not a mere euphemism, but meant to give a different perception of the problem.           Many tools and applications for implicit semantics have been developed for decades and are readily available. Basically all machine learning exploits implicit semantics, namely clustering, concept and rule learning, Hidden Markov Models, Artificial Neural Networks, and others. These techniques supporting implicit semantics are found in early steps towards the Semantic Web, such as clustering in the Vivisimo search engine, as well as in early Semantic Web products, such as metadata extraction on Web Fountain technology, automatic classification, and automatic metadata extraction in Semagix Freedom.

阅读全文(3864) | 回复(0) | 编辑 | 精华