Knowledge Service Based On Bond Information Observation

With the development of the finance big data technology and the acquisition of the investors’ individuation and diversity, finance knowledge service faces the opportunity and challenge. In the process of investing of investor, how to obtain those accurate and real time data, how to estimate the future price and benefit of some bonds, how to select the best bond with minimal risk, how to get the most benefits in the shortest time are some kinds of ineluctable problems that investors may meet. The problems above desiderate the support of technology heterogeneous information processing and data mining. As a consequence, this thesis mainly researches on the method and strategy on observing and mining more evaluable information, and be applied to the service of enterprise bond knowledge.The obtained bond information is the basic of knowledge service. To ensure the accurate and high efficient knowledge service, it firstly needs to get the overall and accurate data, then with the further processing of noise reduction and optimization, transform the collected data to the structured data, provide the data monitor and control for the service processing of recommendation strategy and prediction of tendency.In order to make the investors invest less energy and obtain relatively high returns, we put forward the bond recommendation strategy named ―Same Category Higher Profit‖. From the perspective of bond investors, we analyze and research on the influence in the process of investment in bond yields key features combination, to provide users with more efficient and personalized investment strategy.Bonds trend prediction provides investors with the changing trends of the reference of bond prices and yields. The bond future trends can be predicted based on bond price time series, industry, corporation news and other information with machine learning methods. Integrating multiple factors affecting the characteristics of the bond prices and characteristics in the form, improves the accuracy of the forecast.From the above, observation in the thesis, information from the massive financial data with the NLP and data mining methods can be used to explore more valuable information, to provide users with personalized, diversified and efficient financial knowledge service with bond recommended strategy and trend prediction method.