| As the penetration rate of the Internet continuously grows up, Information generated on the Internet increases explosively. At the same time, the text mining technology provide us a brand new method of investment, which largely differs from the investment framework we are using. With this technology, strategies using market sentiment index are developed.First, this paper compared the analysis framework of the traditional industries and emerging industries. Also we summarized the existing analysis framework of market sentiment.Secondly, we analysis and research of web crawler algorithm and data structure and design a web crawler program capture the internet data, which provides the basic data for indicator construction.Thirdly, we process four proxy variable: quantity of UGC discussion, the number of news reports, capital inflows of the sector and the number of stocks that triggered the raising limit and combine them to the new sentiment indicator. Based on this indicator, we designed an investment strategy.Finally, we use the sentiment indicator of two TMT industries, smart home sector and 3D printing sector, to verify the practical effect. The empirical result prove the effectiveness of the sentiment indicator. |
