Research On Information Advisory Engine System For Financial Investors And Institutions

Information consultation engine system,a kind of vertical search engine,is a system that integrates data of financial institutes,listed companies and local government bonds from various countries on the basis of specific sorting strategy and programming language.It will then present its findings to special user group which mainly contains investors from financial institutes and individual investors.As a result,how to provide a better personalized searching service is the key problem that an application system needs to solve.Therefore,it has great significance in discussing the sorting algorithm and the improved Page Rank algorithm based on user personalized model.The main research fields of this paper are listed below:Firstly,with help of enterprise search engine framework Solr to build up the search engine platform of this system.Study its personalized sorting technology,analyze the user habits of website browsing and calculate the ratio of keywords that are related to users,after which a personalized user model should be able to build based on the data obtained.It is possible to re-sort searching result by this personalized model so as to realize the goal of personalized ranking.It is shown in the study that this re-sort can satisfy customer searching demand better but the efficiency of the searching engine is lowered.On the basis of this,this paper studies two methods to improve the Page Rank algorithm to improve the efficiency of personalized sorting.The first method is Page Rank algorithm which is based on calculating personalized webpage weight.This method makes the website Page Rank value more personalized by analyzing the user logs.The second method is personalized Page Rank algorithm which is based on transaction clustering pattern.It aims to adjust webpage weight by obtaining keywords accessing sequences that are searched by the users.In this way,it can show user personalized feature and present personalized Page Rank algorithm that based on subjective transaction clustering pattern,the main concept of which is to use the user searching keywords and webpage subject to personalized the webpage weight.To test the effectiveness of the methods mentioned,an information consulting search engine for financial investors and financial institutes is developed,which has already been applied in a company business searching platform with success.It is shown by QA testing experiments platform that searching engine based on Solr is better than that based on Endeca;searching result based on user personalized algorithm can better cater user searching demand;the efficiency of personalized Page Rank algorithm based on subjective transaction clustering pattern is significantly higher than Page Rank algorithm based on both personalized webpage weight calculating and transaction clustering pattern.