Research On Intelligent Analyzing For Enterprise Competitive Intelligence Based On Data Mining

With the advent of the era of knowledge economy and the gradually in-depth globalization, information and networking of economy, the market is increasingly competitive, the confrontations among enterprises continually escalate, and the internal and external factors that affect enterprises' business activities are more difficult to predict. Therefore, it is more important for enterprise to make correct competition strategy when they face quickly changed environment besides enlarging the investment in funds, technologies and human resources, in order to survive and develop in such competitive environment. Competitive intelligence, which is taken as the strategic resource for enterprise to keep competitive advantage, and the basis for enterprise to analyze and forecast industry development trend, and to make scientific strategic decisions, is well-known "the forth core competence" of enterprise besides funds, technologies and human resources. It is decision-making knowledge about competitions, competitive environment and corresponding competitive strategy. It can contribute to help and support enterprise organization members to evaluate the key develop trends, track emerging discontinuous changes, grasp the evolution of industry structure, and analyze the capability and tendency of current and potential competitors, providing strong intellectual support and information security for enterprise to keep and develop relative competitive advantage. Intelligence analysis is the crucial link to obtain competitive intelligence.Enterprise competitive intelligence analysis based on data mining absorbs the research achievements form information science, business intelligence, knowledge management and modern competition theory, and supported by many high and new information processing technologies, such as ontology engineering, data warehouse, visualization technology, providing intellectual support for the intelligent mining, analysis, access and innovation of enterprise competitive intelligence and for the decision-making of enterprises. In this paper, the research work mainly includes eight chapters.In chapter1, the basic concepts, features and functions of enterprise competitive intelligence are introduced, the main contents and analysis methods of enterprise competitive intelligent analysis in current knowledge economy environment are analyzed, and the processing of analysis strategy and value increment is also discussed. Moreover, the advantages of the intelligent analysis based on data miningfor enterprise competitive intelligence are expounded.In chapter2, data mining techniques is incorporated into the intelligent analysis of enterprise competitive intelligence. A systematic framework for the intelligent analysis of the enterprise competitive intelligence based on data mining is built, and is elaborated in detail at three levels-theoretical and technology basis, intelligent analysis strategy and menthod, visualization of the intelligent analysis result.In chapter3, this chapter is mainly researched the semantic organization for source data of competitive intelligence by building domain ontology.The construction methods and implementation of software enterprise domain ontology building are studied, and knowledge engineering and thesaurus methods are combined to guide the development of domain ontology; the domain knowledge and conceptual model of Enterprise ontology and TOVE ontology are reused to construct domain ontology framework, and the ontology development tool, Protege, is used to perform the formal codes of software enterprise domain ontology, so that semantic knowledge can be provided for the subsequent data mining and intelligent analysis based on semantic.In chapter4, the methods and algorithms for enterprise competitive intelligence clustering analysis are studied from perspective of clustering mining. Aimed at the problems that the traditional clustering methods and algorithms can't cluster ideal results due to the lack of semantic knowledge, semantic knowledge provided by domain ontology is used to do clustering mining at semantic level to realize the semantic clustering mining and analysis of enterprise competitive intelligence. Onto-kmeans, a k-means semantic clustering mining algorithm based on domain ontology, is designed taking k-means algorithm as the foundation, which is verified by contrast experiments on the WEKA to get more improvement than the traditional k-means algorithm.In chapter5, the methods and algorithms for enterprise competitive intelligence classification analysis are studied from the perspective of classification mining. Aimed at the problems that the traditional classification mining methods and algorithms lack of semantic and need a lot of artificial tagging, semantic knowledge provided by domain ontology is used to do classification mining at semantic level, to realize the semantic classification mining and analysis of enterprise competitive intelligence. Onto-TC, a classification mining algorithm based on domain ontology, which is verified by contrast experiments on the WEKA to be effective.In chapter6, the methods and algorithms for enterprise competitive intelligence association analysis are studied. For the problems of the traditional association mining methods and algorithms, which is resulted from the lack of semantic, such as heavy I/O loads, high overhead and low general obtained rules, semantic knowledge provided by domain ontology is used to realize association mining at semantic level, to implement the semantic association mining and analysis of enterprise competitive intelligence. Onto-Apriori, a semantic association mining algorithm based on domain ontology, is designed taking Apriori algorithm as the foundation. Contrast experiments on the WEKA verifies that the proposed algorithm get more improvement than traditional Apriori algorithm.In chapter7, software enterprises are selected as experimental object. The experimental data is collected from the network information resources. The intelligences, such as factors that influence the competition of software enterprises, are mined and analyzed. Firstly, eight main factors that influence the competition of software enterprises are obtained through semantic clustering analysis. Secondly, taking these factors as classification standards, two classifications are completed by using semantic classification analysis. Finally, semantic association mining is applied to get the semantic associations between these chief elements and specific influential factors, which are beneficial for software to the cultivation of core competitiveness and the prediction of market risks.In chapter8, a systematic summarization for the main contents of this paper is given, and a prospection is proposed for future work. The former is done with research meaning and contents. The latter is focus on the future research directions of enterprise competitive intelligence intelligent analysis based on data mining research, which include the mining and analysis of intelligence with complex data types, visualized intelligence, dynamic intelligence and knowledge centered big data, etc.This paper is sponsored by National Natural Science Foundation "Intelligent Analyzing Model and Method for Enterprise Competitive Intelligence"(Project ID:71073121), Ministry of Education for the Major Research Base of Human Social Science "Knowledge Mining Technology Based on Intelligent Information Processing and its Application"(Project ID:08JJD870225), and Ministry of Education for the Doctor Academic Award "Business Intelligence Analyzing Method Based on Data Mining"(Project ID:5052012104001).