Research On The Acquirement Of Enterprise Competitive Intelligence In The Web

Competitive intelligence has been the fourth core competitiveness at the age of knowledge economics and informatization. With the rapid development of Web technologies, how to effectively and efficiently acquire enterprise competitive intelligence from the Web has been a new topic in the research on competitive intelligence. However, previous approaches were restricted on the Web-page-driven ways and focused on collecting Web pages and perform textual Web search. According to those approaches, competitive intelligence were mainly collected and analyzed by means of Web search engines or text mining tools. As previous ways did not conduct deeply understanding and extraction on Web information, there will consequently be a big gap bwteen the extracted results and users’needs. Such situations have hindered the further development of the theories and applications of enterprise competitive intelligence in the Web environment.In this dissertation, we aim at satisfying the requirements of collecting enterprise competitive intelligence in the Web environment and solving the critical issues existing in this procedure. In particular, we concentrate on the representation model of the enterprise competitive intelligence in the Web environment, as well as the methods to acquire enterprise competitive intelligence in the Web with respect to different viewpoints. As Web can be regarded as a platform of information resource which involves Web pages, Web sites, and Web users, we conduct our research on competitive intelligence acquirement based on three viewpoints, i.e., the Web-page-based viewpoint, the Web-site-based viewpoint, and the Web-log-based viewpoint. Our algorithms are expected to present a systematic framework for the acquirement of enterprise competitive intelligence in the Web, and thus to form the foundation for the future research on Web-oriented enterprise competitive intelligence and applications.In general, the contributions of the dissertation can be summarized as follows:(1) We study the semantics of the enterprise competitive intelligence in the Web environment, and propose an entity-based representation model for Web-based enterprise competitive intelligence. Based on this model, we develop a framework to extract competitive intelligence from the Web. Our framework is founded on entity recognition and relation extraction, and provides a fundamental solution to the extraction of Web-based competitive intelligence.(2) We study the issues on acquiring enterprise competitive intelligence from Web pages, and present a framework to describe the business relations of competitors. After that, a new algorithm to extract company acquirement relations from Web pages is proposed, which is based on the tense labeling for sentences in the pages. The experimental results demonstrate its effectiveness.(3) We research the issues on acquiring enterprise competitive intelligence from the Web-site viewpoint, and propose a process as well as an example to analyze competitors’intelligence by utilizing Web sites information. This study can offer new insights to the acquirement and analysis of Web-based enterprise competitive intelligence.(4) We explore the extraction of competitors’intelligence from the logs of Web users, and present a process model to perform competitor analysis by suing the behavior logs of Web users. We take the electronic business area as an example and conduct comparable analysis on the typical electronic business companies on the basis of real behavior logs from Internet users. This research provides a new way to analyze competitor intelligence, and is of referential values to the acquirement and analysis of Web-based enterprise competitive intelligence.