Description: This paper presents an approach to mining domain-dependent ontologies using term extraction and relationship discovery technology.There are two main innovations in the approach. One is extracting terms using log-likelihood ratio, which is based on the contrastive probabilityofterm occurrence in domain corpus and background corpus. The other is fusing together information from multiple knowledge sources as evidencesfor discovering particular semantic relationships among terms. In the experiment, traditional k-mediods algorithm is improved for multi-levelclustering. The approach to produce an ontology for the domain of computer science is applied and promising results are obtained.
File list (Check if you may need any files):
072282.pdf