无线城市信息库的用户兴趣模型的设计与实现

发布时间：2018-04-25 18:46

本文选题：无线城市 + 信息库　；参考：《北京邮电大学》2013年硕士论文

【摘要】：作为智慧城市在中国的应用,无线城市项目的提出旨在解决市民生活信息化问题。无线城市的核心是如何将城市中大量的数据进行收集、整合和处理,并将处理之后的数据转化为信息提供给应用使用。与此同时,数据的处理离不开无线城市最终的受众即用户的参与,如何避免信息过载问题的出现,将数据转化为有实际意义和用户感兴趣的信息是无线城市亟待解决的问题。作为无线城市的核心模块,信息库系统运用知识工程和知识库的概念与方法,对无线城市中涉及的海量数据进行统一的模型存储、映射和关联分析。信息库的功能是实现无序数据的规则化,并发掘信息之间的关联性,为最大限度的聚合信息和满足用户的个性化需求而服务。信息库负责实现的是从数据采集、数据抽取、数据分析、数据建模、数据加工到用户推荐这一系列流程。信息库中的核心概念是信息与用户。用户兴趣模型研究的正是如何在用户最少的参与下,分析用户的行为路径,发掘用户的兴趣所在。信息库系统中的用户兴趣模型,是在传统用户兴趣模型基础上,针对无线城市海量数据和不同类别应用而提出的。它的目标是对无线城市中海量无规律的数据进行处理,并分析其中与用户相关的信息,将所有与用户相关联的信息进行加工整理,从而实现用户兴趣的模型化记录和表达。同时,基于用户兴趣模型实现包括推荐系统和搜索引擎在内的服务功能,为用户推荐其感兴趣的结果,验证模型的可行性和效率,从而更好地修正和更新兴趣模型。本文的主要工作可以归纳如下： (1)分析无线城市的实际需求和应用场景,总结归纳无线城市数据特性,提出无线城市信息库的设计思路和实现方案。 (2)分析传统用户兴趣模型的优缺点,结合无线城市信息库的特质,提出无线城市信息库用户兴趣模型的处理流程。 (3)按照数学建模的思想,结合统计学、数据挖掘、自然语言处理和图论的知识与算法,设计用户信息相关度模型、用户内容相关度模型和用户行为相关度模型,并实现模型融合技术。 (4)实现用户兴趣模型的自动化更新,结合搜索引擎和推荐系统的功能,实现基于用户兴趣模型的服务,并解决冷启动问题。 (5)讨论无线城市信息库用户兴趣模型的通用性和推广方法,为其它系统采用此用户兴趣模型提出建议。 (6)结合推荐系统的测试方法和评测指标,将本系统提出的方法与传统用户兴趣模型的方法进行对比和结果分析,验证效果。
[Abstract]:As the application of Intelligent City in China, wireless city project is put forward to solve the problem of informatization of citizen life. The core of wireless city is how to collect, integrate and process a large amount of data in the city, and convert the processed data into information for application. At the same time, the processing of data can not be separated from the wireless. The final audience of the city, the participation of users, how to avoid the emergence of information overload, and transform the data into practical and user interested information is an urgent problem to be solved in the wireless city.
As the core module of the wireless city, the information base system uses the concepts and methods of knowledge engineering and knowledge base to carry out unified model storage, mapping and association analysis of massive data involved in the wireless city. The function of the information base is to realize the regularization of disordered data, the association of the data and the maximum aggregation. Information and service to meet the personalized needs of the user. The information base is responsible for implementing the series of processes from data collection, data extraction, data analysis, data modeling, data processing to user recommendation.
The core concept in the information base is information and users. The research on user interest model is how to analyze the user's behavior path and discover the user's interest under the least user participation. The user interest model in the information base system is based on the traditional user interest model, which aims at the massive data of the wireless city and the different kind. The goal is to deal with the irregular data in the wireless city, analyze the information related to the user, and process all the information associated with the user, so as to realize the model recording and expression of the user's interest, and the implementation of the user interest model, including the recommendation system and the search. The service function of the engine is used to recommend the result of interest to the user, verify the feasibility and efficiency of the model, so as to better modify and update the interest model.
The main work of this article can be summed up as follows:
(1) analyze the actual needs and application scenarios of wireless cities, summarize the characteristics of wireless city data, and put forward the design idea and implementation plan of wireless city information base.
(2) analyzing the advantages and disadvantages of the traditional user interest model, and combining the characteristics of the wireless city information base, we propose the processing flow of the user interest model of the wireless city information base.
(3) according to the idea of mathematical modeling, combining statistics, data mining, knowledge and algorithms of Natural Language Processing and graph theory, we design user information correlation model, user content correlation model and user behavior correlation model, and realize model fusion technology.
(4) the automatic update of the user interest model, the function of the search engine and the recommendation system, the service based on the user interest model, and the cold start problem.
(5) discuss the universality and promotion method of user interest model of wireless city information base, and provide suggestions for other systems to adopt this user interest model.
(6) according to the test method and evaluation index of the recommendation system, the method of this system is compared with the traditional user interest model, and the results are analyzed and the results are verified.

【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP391.3

【参考文献】