基于众包UGC的交通用户分类推荐模型

发布时间：2019-06-05 23:43

【摘要】：在WEB 2.0时代最近的10年里,数量呈几何倍数增长的互联网网民以及随科技发展而不断变化的新兴网络交互方式,使得大量的信息得以在众多互联网平台中产生并传播。智能交通信息平台作为WEB 2.0平台时代的“后起之秀”,目前仍然依靠大范围部署传感器等硬件,作为交通信息的主要采集方式,不可避免地暴露出了依靠设备采集数据的局限性,并造成了对于平台存在的海量UGC数据的浪费。本文针对这一背景及国内外研究现状,提出了一种基于众包UGC的交通用户分类推荐模型。以模型为基础,本文主要进行了以下三方面的研究。1、交通用户行为特性与众包UGC的关系,且该关系所反映的用户属性的相关研究。2、交通用户分类推荐模型依赖的判别标准及相关算法研究。3、以产生信息为主的智能交通信息平台的仿真。本文以笛卡尔坐标系为骨架的智能交通信息平台作为仿真基础,结合WEB 2.0时代的平台特征、众包UGC特点、用户行为特性等,借助质量控制策略以及合作博弈中的Shapley权力指标、图论相关知识、最短路径算法、路线匹配的动态规划算法、分词工具等,将众包UGC与用户分类的推荐方案相结合,并进一步通过数据验证了模型在交通信息平台中的有效性及可行性。基于众包UGC的交通用户分类推荐模型得到的结果为用户提供了相对实时的交通信息推荐、好友推荐等,方便用户获取其所真正关心的信息,并做到了“取之于用户,用之于用户”。而这便是本课题的终极意义。
[Abstract]:In the last 10 years of the WEB 2.0 era, the number of Internet netizens with geometric multiples and the new network interaction with the development of science and technology have enabled a large number of information to be generated and disseminated in many Internet platforms. As a "rising star" in the era of WEB 2.0 platform, Intelligent Transportation Information platform still relies on large-scale deployment of sensors and other hardware, as the main way to collect traffic information. Inevitably, the limitation of relying on equipment to collect data is exposed, and the waste of massive UGC data existing on the platform is caused. In view of this background and the research status at home and abroad, this paper proposes a traffic user classification and recommendation model based on crowdsourcing UGC. Based on the model, this paper mainly studies the following three aspects. 1. The relationship between traffic user behavior characteristics and crowdsourcing UGC, and the related research of user attributes reflected by this relationship. Research on discriminant criteria and related algorithms of traffic user classification recommendation model dependence. 3. Simulation of intelligent transportation information platform based on information generation. In this paper, the intelligent traffic information platform based on Cartesian coordinate system is used as the simulation foundation, combined with the platform characteristics of WEB 2.0 era, the characteristics of crowdsourcing UGC, the characteristics of user behavior and so on. With the help of quality control strategy and Shapley power index in cooperative game, graph theory related knowledge, shortest path algorithm, dynamic programming algorithm of route matching, word segmentation tool and so on, the crowdsourcing UGC is combined with the recommended scheme of user classification. The validity and feasibility of the model in traffic information platform are verified by the data. The results of the traffic user classification recommendation model based on crowdsourcing UGC provide users with relatively real-time traffic information recommendation, friend recommendation and so on, which is convenient for users to obtain the information they really care about, and achieves "taking it from users." Use it for users. And this is the ultimate significance of this topic.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.3

【参考文献】