Static and Dynamic Data Collection and Data Mining for Small Business Domain Knowledge

Abstract

In the modern era everyone using web for analysis and analytics of large amount of data. In addition, most of the web pages did not contain semantic information until recently, raising difficulties for mechanical research. As a way to remedy these poor online characteristics, the Semantic Web has arisen. In this report, with the aid of Twitter data mining, we aimed to build up a business knowledge domain to provide useful information for small business owners for their marketing campaigns or dynamic QA systems for their business recommendation services. According to the Semantic Web principle, the knowledge base was constructed. First, web crawling from various web sources, including social media, is required to build the information domain. The crawled data, however, is usually accessible informally and has no semantic knowledge. In order to capture useful information from them and generate formal knowledge for the knowledge domain, we developed text mining techniques. Sentimental analysis based on Stanford NLP also runs with this framework for accuracy.