Intelligent information mining from veterinary clinical records and open source repository

This paper reports an implementation of an intelligent mining approach from veterinary clinical records and an external source of information. The system retrieves information from a local veterinary clinical database and then complements this information with related records from an external source, OAIster. It utilizes text-mining, Web service technologies and domain knowledge, in order to extract keywords, to retrieve related records from an external source, and to filter the extracted keywords list. This study meets a practical challenge encountered at the School of Veterinary and Biomedical Sciences at Murdoch University. The results indicate that the system can be used to increase the limited knowledge within a local source by complementing it with related records from an external source. Moreover, the system also reduces information overload by only retrieving a set of related information from an external source. Finally, domain knowledge can be used to filter the extracted keywords, in this case, selected medical keywords from the extracted keyword list.