Research On Competitor Identification Based On Social Network Analysis

Competitive intelligence has undoubtedly been "the fourth core competencies" for the enterprises.With the explosive in web information, how to effectively obtain competitive intelligence information from the web is a challenge,the difficulty of the work increasing.The primary task of competitive intelligence is competitor analysis, competitor analysis,which is the first step to identify competitors, therefore, accurately identify competitors in the enterprise competitive intelligence work is particularly important.Social Network Analysis providing a sociological perspective, brings a new way to identify competitors. This dissertation analyzes the web link structure, combined with social network analysis, researches the identification of competitors, and it aims to enhance the objectivity of identify competitors results, especially potential competitors, which provides reference for future research.At the beginning, the author found there are limited problems in the most study after outlined the analysis based on social network research in recognition competitors. First,when conducting web link structure analysis, most scholars only consider the number of links between enterprises nodes, the result is not objective enough; Second, most of the research didn’t the competitors for future. To resolve the problems, this thesis proposes the enterprise’s core competitiveness evaluation index system based on value chain analysis,aiming to estimate the node website quality with quantitative analysis and comparative study verified that the index system can improve recognition results competitors objectivity.Next, propose an improved social network links prediction algorithm,,and the improved algorithm can be verified by experiment that accuracy and time complexity predictions could effective unity. Finally, based on the index system,combined with the improved algorithm, using the web link structure data, empirical research in the field of competitors to systematically demonstrating competitors identification process, at this stage of the competition for the new energy automotive industry recognition results described in detail.The results show that, in the social network analysis, consider the quality of the node enterprises in the link prediction algorithm,to some extent, can improve the phenomenon of the objectivity of competitors results not good enough, but also provides a reference for identifying potential competitors.