Automated Discovery of Small Business Domain Knowledge Using Web Crawling and Data Mining

It has become an era where everything is on the web with ever more chances of data utilization on the web. Still, there are obstacles to make the use of the web efficiently. With too much information, Internet users have often come across information that are not relevant for their use. On top of that, until recently, most of web content have not contained semantic information, posing difficulties for mechanical analysis. The Semantic Web emerged as a way to tackle those poor qualities of the web. Adopting formal languages such as RDF or OWL, the semantic web has made the Internet become more highly available for computer-based analysis. In this study, what we aimed at is building a small business knowledge base to provide useful information for small business owners for their marketing strategies or dynamic QA systems for their restaurant recommendation services. The knowledge base was built according to the concept of the Semantic Web. To build the knowledge base, first, it is needed to conduct web crawling from different web sources including social media. However, the crawled data typically come in informal and do not have any semantic information. So we devised text mining techniques to catch useful information from them and generate formal knowledge for the knowledge base.