Knowledge base Internet search for construction and real estate information system

Abstract

Information is at the centre of the construction and real estate management. In the past, construction professionals have to go through tedious operations to find the relevant information in need. This thesis presents a feasible solution to solve the problem stated. Nowadays information is increasingly available in web format that can be reviewed via browsers. However surfing the Internet is perceived as inefficient and ineffective for professional use. Search engines are used instead to find the required information based on user's specification (often in words). Based on techniques established in pattern recognition and experience of using knowledge bases in intelligent applications, the knowledge base approach has been proposed to significantly increase the usefulness and effectiveness of search engines used in the area of professional applications. The knowledge base approach has the advantages of being more relevant to user and adaptable to shift in environmental parameters. A prototype system was built to illustrate the knowledge base Internet search approach. It was observed that feature vectors had been successfully used in searching for the underlying texts. Search results were accumulated and processed for new knowledge. The acquired new knowledge was found to be useful in enhancing the relevancy of subsequent search process. Practically the knowledge base approach can be used unattended to automatically generate new and relevant knowledge for professional users.