证券领域个性化推荐平台

发布时间：2018-01-12 06:11

本文关键词：证券领域个性化推荐平台　出处：《电子科技大学》2014年硕士论文　论文类型：学位论文

【摘要】：证券是我国金融行业的一个重要领域,它不仅关系到我国市场化的进展,还关系到企业改制的顺利进行,但是,随着股票发行的核准制度、证券交易浮动的佣金制度、强化上市公司的信息披露、证券对国际市场的逐步开发、大力推行证券投资基金等措施的实施,证券领域从过去的垄断暴利行业转化成如今的开发竞争性的微利行业,证券领域的发展受到了严峻的挑战。在这种严峻的挑战下,证券营销就变得空前的重要,随着信息技术的高速发展与应用,利用信息化技术来解决证券营销问题已经成为当前的一个研究热点。本文力图从信息化技术理论和实际出发,提出一种证券领域的个性化推荐框架平台和两种新的证券营销策略。在证券领域,随着证券业务的飞速发展、证券产品的类型越来越多和证券客户历史交易信息的急剧增长,用户需要花费大量的时间与精力才能找到自己所需要的信息,这种浏览大量无关信息的过程无疑会使淹没在信息过载问题中的用户不断流失,给证券公司带来恶劣的后果,在这种情况下,证券营销的推荐需求特别是个性化推荐的要求越来越强烈。为了改善证券领域中传统的大众(无差别)推荐,本文提出了一种证券领域的个性化推荐框架平台和两种更为有效的个性化推荐设计方法。(1)针对证券信息的个性化推荐平台:本推荐平台将系统分成13个独立且相互联系的子模块,各模块之间通过相互调用来完成最终的推荐任务,还给出了相应的数据流及应急设计的方案。(2)基于用户行为的个性化推荐方法:通过分析用户在股票、理财产品、资讯等证券信息的浏览、购买、订阅等行为,设计并实现个性化推荐列表,并从真实数据的实验结果中可以看出此推荐方法是有效的。(3)基于文本分析的个性化推荐方法:在证券领域有很多信息都是以文本的形式存在,然而它们却有着非比寻常的意义,基于文本分析的个性化推荐方法充分的利用文本内容的天然优势为用户提供个性化的推荐列表,此方法也可很好的解决行为推荐方法中的冷启动问题,并真实数据的实验结果中可以看出此推荐方法是有效的。
[Abstract]:Security is an important component of China's financial industry, it is not only related to the progress of our country market, but also related to enterprise restructuring smoothly, however, as the stock issuance approval system, the securities and Exchange Commission floating system, strengthen information disclosure of listed companies, securities development gradually to the international market, the implementation of vigorously promote the securities investment fund and other measures, the field of securities conversion from the past monopoly profiteering industry development competition of today's low profit industry, development of securities market has been a severe challenge. In this challenge, the securities marketing becomes ever more important, with the rapid development of information technology and Application, to solve the the problem of securities marketing and use of information technology has become a research hotspot. This paper starts from the theory and practice of information technology, put forward a kind of personalized domain security The recommended framework platform and two new types of securities marketing strategy. In the field of securities, along with the rapid development of the securities business, the rapid growth of securities products more and more types and securities customer transaction information, users need to spend a lot of time and effort to find the information they need, the process of browsing a large number of irrelevant information will undoubtedly the submerged in the problem of information overload in continual loss of customers, to bring bad consequences to the securities company, in this case, the recommended demand of securities marketing especially more and more strong requirements of personalized recommendation. In order to improve the traditional public securities in the field (no difference) recommendation, this paper proposes a personalized stock field the recommended framework platform and two kind of more effective personalized recommendation method. (1) according to the securities information recommendation platform: this recommendation platform system is divided into 13 independent and interrelated sub modules, each tune used to complete the final recommended tasks through among modules, data flow and the corresponding emergency design scheme is given. (2) the personalized recommendation method based on user behavior: through the analysis of the user in the stock, financial products, information and other securities information browsing. The purchase, subscription and other acts, the design and implementation of personalized recommendation list, and experimental results from real data can be seen in the recommended method is effective. (3) the personalized recommendation method based on text analysis: in the field of securities have a lot of information are in the form of text, but they are of great significance the use of the natural advantages of text content, text analysis and personalized recommendation method provide personalized recommendation list for users based on this method can resolve the behavior of good recommendation method in cold start The problem, and the experimental results of real data, can be seen that this recommendation method is effective.

【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：F832.51;TP391.3

【相似文献】