【作者】 李海强；

【导师】 吕廷杰；

【作者基本信息】 北京邮电大学 ， 管理科学与工程， 2009， 博士

【摘要】 信息通讯社会正朝着一个环境感知泛在网络AUN(AmbientUbiquity Network)的生态环境迈进,网络就如同空气和水一样,自然而深刻地融入了人们的日常生活及工作中。网络将不再被动地满足用户需求,而是主动感知用户场景的变化并进行信息交互,通过分析人的个性化需求主动提供服务;过去对移动业务的研究主要集中在对业务本身的研究,很少考虑移动用户对业务的喜好,以及用户所处的位置和时间,对下一代移动网络来讲,这是不足够的;并且无线网络的发展允许移动用户在任何时间任何地点都能得到感兴趣的移动业务,网络能主动感知移动用户场景的变化,从而提供将要需要的业务及信息;特别是随着无线技术的发展,例如:GPS和RFID技术的广泛应用,移动情景感知业务将是移动运营商未来发展的重点业务。因此,如何帮助用户获取感兴趣的业务信息是一件急需解决的问题,这就需要对移动用户的行为模式进行深入研究,使网络可以智能感知到用户的序列移动接入模式,并在相应的位置和时间智能地为用户提供所需要的业务,而移动用户的行为模式与用户所处的场景和相应的业务需求有密切的关系,序列移动接入模式也成为分析情景感知业务的一个重要研究方向。为此,本文基于中国移动的手机地图业务,从情景感知的移动用户行为模式角度出发,针对用户的序列移动接入模式挖掘、个人隐私泄露问题和移动接入模式的预测三个方面进行了深入分析,提出了合理的理论解决方案,并挖掘出相应的关键技术问题进行了深入研究。(1)用户的序列移动接入模式挖掘研究移动用户的业务接入模式需要充分考虑移动用户所处的位置、时间和业务需求三个属性。目前,国内外的参考文献主要基于位置和业务需求两个属性进行分析。其中,有一些文献也从分段的时间角度考虑,结合用户所处的位置和业务需求设计了序列移动接入模式的挖掘算法,这种算法可以大致的挖掘出分段时间区域内的频繁时间序列移动接入模式。但是,这种人为对时间进行分段,很容易将一些频繁序列移动接入模式分成几段,不利于对整个移动业务接入模式的分析。因此,国内外对移动用户序列接入模式的挖掘研究尚处于起步阶段,还有很多工作需要做。(2)基于情景感知业务的个人隐私泄露问题研究隐私泄露已经不是一个新的问题,但随着网络技术的发展与移动商务的兴起,“隐私泄露”问题也逐渐被放大;特别是在新的移动商务环境中,隐私泄露与保护的问题已经成为影响移动商务未来发展趋向的重要议题之一。在情景感知为我们未来的生活带来无限便利的同时,也使得用户对个人隐私的安全产生了担忧,使得用户个人隐私面临巨大的泄露危险。因此,在为用户提供便捷、有效服务的前提条件下,如何限制移动用户的隐私泄露?也就是如何将个人隐私的泄露限制在无害的范围内成为限制情景感知业务未来发展的重点。(3)序列移动接入模式的预测研究面对海量数据,数据挖掘技术可以帮助我们获取有用的信息。目前,国内外关于情景感知方面的信息挖掘与预测研究相对较少,相关的研究机构大多集中在Web挖掘预测方面的研究,然而,两者具有很多相似性。Web推荐的任务是将当前用户的会话与挖掘阶段分析出的模式相匹配,并为当前用户推荐一个包括商品、超级连接等的项目集;而序列移动接入模式的挖掘预测则是根据运营商数据库中关于移动用户所发生的历史业务信息进行挖掘和预测。两者都参照了用户的历史信息,对用户将来的行为进行预测。但是,两者在参考用户属性方面有很大的差异性,Web推荐主要考虑用户会话、URL的连接距离等因素,而序列移动接入模式的预测主要考虑用户所处的位置、连续时间和用户的业务需求等相关因素。因此,基于上述的相似性和差异性,两者需要设计相应的不同预测算法去解决各自的问题。如何去设计序列移动接入模式的预测算法是本文的重点研究内容。总之,国内外对用户的序列移动接入模式的预测研究还处于起步阶段,如何考虑时间、位置和用户需求等多维属性来对其进行研究是一个重要的研究方向。更多还原

【Abstract】 The rapid advance of ubiquitous mobile network technologies enables the provision of rich kinds of mobile services based on context awareness for mobile users. The future wireless network, as air and water, penetrate our daily life and work, which will not be passive to satisfy mobile end user’s requirements, but initiate actively to aware end users context’s shift and analyze their personal requirement, and then exchange their information. The past studies on mobile services focused mainly on the provision of anytime and anywhere services. However, this is insufficient for the future mobile internet systems. The context characteristics, such as the user’s location, the user’s preference, the user’s service request time and the user’s historical behavior, should also be considered in order to provide the users with context-aware services that will benefit the users.The development of future mobile internet has allowed the mobile users to request various kinds of services by mobile devices at anytime and anywhere. Helping the users obtain needed information effectively is an important issue in the future mobile internet systems. Discovery and analysis of mobile user’s diverse behavior can highly benefit the enhancements on mobile internet system performance and quality of services. Obviously, the mobile user’s behavior patterns, in which the location and the service are inherently coexistent, become more complex than those of the traditional web systems. Therefore, this paper makes a deep research and analysis on sequence mobile access pattern mining method, privacy limiting method in context awareness data mining and prediction algorithm for mobile access pattern, and presents the reasonable theoretical solutions on the key technical issues:(1) Sequence mobile access pattern mining research for mobile user;The mobile user’s diverse behavior patterns are usually associated with the user’s location, the user’s service request and the user’s service time. Recently, some studies have been done on mining sequence mobile access pattern with the user’s location and service request considered. Some research teams also simply consider the user’s service time characteristic, they divide one day, 24 hours, into several section, Thus, the mobile access pattern for every time section was analyzed, so that some key sequence mobile access patterns are easy to be divided by the time segment. But, no studies have focused on continuous time sequence mobile access pattern so far, and the relative research is in the beginning phase, and there are more relative work need to do. (2) Personal privacy leakage research based on context awareness service;With the development of mobile network technologies and mobile commerce, personal privacy leakage had not been a new issue. Especially, privacy leakage and protection issues had affected the future development of new mobile commerce. Context awareness services enable user’s convenient life, whereas, privacy leakage issues from context awareness service will be worried about by mobile users simultaneously. Therefore, how to limit mobile user’s privacy leakage will be the future context awareness service’s key research content.(3) Prediction research on sequence mobile access pattern;Recently, few studies have been done on mobile access pattern prediction based on context awareness, and the relative research teams mainly focused on web data mining and prediction. However, both of them have some similarities, that is, all of them predict mobile users’ future behavior based on their historical behavior information. However, there are also some dissimilarities between mobile access pattern prediction and web data prediction. Web data prediction mainly considers below factors, such as user session, URL linkage etc., while mobile access pattern prediction mainly involves below factors, such as mobile user’s location, mobile service time and mobile user’s service request etc.. Therefore, they need to design different prediction solutions to predict future mobile user behavior pattern based on their similarities and dissimilarities. How to design sequence mobile access pattern prediction algorithm is considered as the paper’s key research content.In conclusion, sequence mobile access pattern prediction research based on context awareness is just in a beginning stage for the international research teams, and thus how to design mobile access pattern’s multi-dimension model including the user’s location, the user’s service request and service time considered will be an important research area.更多还原