Feature Extraction And Situation Analysis For Vertical Industry

For over 30 years of reform and opening up, the economic development of our country have made remarkable achievements, all works of life got unprecedented development. However, when the economic development step into the deep water, many industries’ development meets bottlenecks and their development prospect is not very clear, which not only affects the investments decision of investors, but also affects the government’s macro control of the country economic. Therefore, understanding industry development present situation and being clear about industry development in the future, have become a reasonably important task to be urgently solved. This paper is written for vertical industry information collection and situation analysis, which can help people understand the industry development history, the current development status and to some extents reflects the industry development trend in the future.Firstly, in this paper, the related theories about finding out characteristics and industry analysis were reviewed. This paper uses the method that is based on industry feature word recognition, in briefly, this method includes three steps: the determination of candidate industry feature words, the iterative extension of candidate industry feature words, the screening of industry feature words. Then, according to the industry feature words, being based on two dimensions of time and space, using the relative weight and its change, this paper analyses the subject evolution of industry development and regional industry development situation, at the same time, this paper display the results visually. Lastly, this paper uses the real estate industry as an example, collects industry information based on apache program----Nutch, recognizes the industry feature words using the feature extraction method, then analyses specially both the Beijing and Shanghai real estate industry development status based on the industry feature words. The experiment result shows that the method that this paper proposes can recognizes the industry feature words efficiently and accurately, describes industry development history, present situation and the future development trend clearly.