26th International Association for Management of Technology Conference - IAMOT 2017

A large amount of data produced in the Internet every day has attributed to the rapid development of mobile Internet technology, the wide application of E-commerce, the popularity of intelligent mobile equipment and so on. Meanwhile, these data involve a mess of unstructured data, such as text data and multimedia data. However, the problem to be solved at the present is how to acquire potential and valuable information by making full use of unstructured data analysis based on Internet technology. This paper targets consumer demand and finds several problems existing in current demand analysis methods, which include inaccurate demand expression, belated demand source, high demand analysis cost and limited demand-acquisition approach. In the digital environment, consumers can express their comments on products and hopes for product quality. Therefore, those existing problems can be solved by discovering consumer demand from those online comments. At the same time, in terms of the rapid development of text mining methods like crawler, natural language processing, text categorization and cluster. Based on the above information, this paper makes an analysis of consumer demand by adopting text mining methods. First and foremost, this paper starts with theory of consumer demand analysis, analyzing source of consumer demand and its five characteristics which consist of dynamics, complexity, hiding property, affective reaction and fuzzification. It combines cyber-expressions of consumer demand and characteristics of text mining, thus discovering several advantages that text mining has when used in the field of consumer demand analysis. Those advantages include reliability, rapidity, low cost, durative and real-time. Based on specific process of text mining, namely, data acquisition, processing and visualization, it also put forward that consumer demand analysis can be divided into demand acquisition, identification and expression, thus reaching hierarchical structure model of consumer demand analysis based on text mining. Then by compares and studies of text mining as well as hierarchical structure mode of consumer demand analysis, it considers that the acquisition of consumer demand information facing online comments can be finished by crawler, and vector space model and clustering analysis algorithm can be better used to identify and express consumer demand, thus putting forward double-spiral structure of consumer demand analysis. At last, this paper recounts the specific flow of consumer demand analysis based on text mining. Research findings of this paper consist in the conclusion of common steps of consumer demand analysis and division into demand acquisition, identification and expression, thus promoting hierarchical structure model of consumer demand analysis and double-spiral structure of consumer demand analysis. Besides, it also analyzes empirically and proves validity and feasibility of consumer demand analysis based on text mining. By means of deep analysis of consumer demand and research of text mining, research significance of this paper lies in providing a new consumer demand analysis method for enterprises, therefore, in the digital era, enterprises can increase acquisition rate of consumer demand in virtue of text mining methods, so that enterprises can rapidly identify consumer demand and help enterprises accurately target the real demand of consumer, discover potential demand of target consumer and promote more suitable products for consumer demand.