Personalized Recommendations Platform For Securities

Security is an important area in the financial industry. It is not only related to the progress of the market in China, but also related to the smooth progress of enterprise restricting. However, with the measures implemented, which are approval system for the issue of shares, securities trading floating commission system, strengthen information disclosure of listed companies, the gradual development of the international securities markets, promoting the implementation of measures such as securities investment funds, security field from the monopoly profits industry in the past turns into today’s open competitive low- profit industry, and its development subjects to serious challenges. Facing to the serious challenges, the security marketing becomes ever important. With the rapid development and application of information technology, the use of information technology to solve the problems of security marketing has become a research hot spot. From the point of information technology theory and reality application, the thesis proposes a platform of personalized recommendation framework and two new marketing strategies of security.In the field of security, with the rapid development of the security market, more and more types of security products and the sharp increase in the history information of the security trading, users need to spend a lot of time and effort to find the information which they want. However, the progress of browsing much irrelevant information will upset the users and drives them away from security productions, which may cause great loss for security companies. In this case, the request for recommendation in the security marketing is more and more intense, especially personalized recommendation.In order to improve the traditional mass(undifferentiated) recommendation, the thesis proposes a platform of personalized recommendation framework and two new marketing strategies of security.1) Personalized recommendation platform for security information: The recommended platform system is divided into 13 independent and interrelated sub-modules. These modules are by calling each other to complete the final task. The thesis also gives the corresponding design data flow and contingency plan.2) A personalized recommendation method based on users’ behavior: By analyzing users’ behavior such as browsing, purchasing and subscribing securities information including stocks, financial products, and recent news, the thesis develops and implements a personalized list of recommendations. Experiments are conducted and the test results can prove the efficiency of the recommendation method.3) A personalized recommendation method based on text analysis: In the field of security, there is a lot of information is in the form of text, and they are very significant. The personalized recommendation method based on text analysis makes full use of the natural advantage of text to provide users with a personalized list of recommendation. This method can be a good solution to cold-start problems of behavior recommendation method. And the recommendation method is effective from the experimental results.