Tag Extraction Of Stock Code Based On Mapreduce And Assocatiated Recommendation Of Financial News

As the financial planning gradually entered people’s life and the further opening of the securities market, stock trading as a low threshold security which became the preferred security transactions of the broad masses of people, but large numbers of stocks has people spend a lot of time to find the information they need.In order to improve the efficiency of the broad masses of investors browsing the stock code, this paper proposes a method of extracting tag for each stock code, tag information obtained by extraction meet the urgent needs of the broad masses of investors quickly understanding the relevant information of each stock code. Tag extraction method of this paper is based on the assumption: a really important tag of a stock code appears in any existed and associated financial news, thus the tag extraction process is divided into two steps:first is to extract, namely single document keyword extraction; followed by calculation, that is to calculate the probability calculation for extracted keywords to obtain the tags. In order to simultaneously consider the connection between the word and word inside the same news and the connection between the word and word among different news of the same stock code, this paper proposes and implements TextRank-WBayesian and proves the feasibility and effectiveness of the method through the analysis of experimental results. In the information age, the daily production rate of information is accelerating, while the stock market is also unpredictable and ever-changing, it needs to do the tag extraction every day, in order to improve the program’s scalability and efficiency, the tag extraction program proposed by this paper will be based on Hadoop MapReduce which is an open source computing framework.In order to make the broad masses of investors get a better investment orientation, this paper proposes an associated recommendation algorithm for financial news of the stock code, so that the broad masses of investors can timely have access to their own stocks’ related information as related industries, competitive corporation and partner corporation, thus the investors can better seize investment opportunities and get more revenue. By sampling analysis, the experimental results prove that the associated recommendation can better reflect the trend of the stock market.In order to evaluate the tag extraction of the stock code and associated recommendation of financial news more objectively, in the future, it needs to establish a common tag evaluation set and general evaluation algorithm.