业务上下文的处理机制及其预测理论、关键技术研究
|
摘要
智能化、个性化是新一代网络业务发展的趋势。未来的业务将需要跨越多种承载网络、跨越多个运营商域,具有普遍查询访问,按需组合,上下文处理,移动无缝应用的能力,从而形成一个以用户为中心UC(User-Centric)的人性化、智能化的业务环境。业务上下文信息处理机制是为用户提供智能化、个性化业务的前提基础,涉及到网络、终端和用户环境的同源同质或同源异质多种上下文信息的融合和处理。本文主要针对新一代网络的业务应用层面展开深入研究,主要包括未来的业务上下文信息的处理机制和上下文预测理论以及关键技术研究。
(1)首先,针对目前业务上下文信息处理机制还不完善,缺乏可扩展的通用的上下文信息处理支撑平台,建立了具有普遍意义、可扩展和分层的多源异质或同质的业务上下文信息的处理机制架构SCIPMA(Service Context Information Processing Mechanism Architecture),确立了业务相关的上下文信息的范畴和层次框架,以支持业务的智能化处理机制,为用户提供个性化业务。在此基础上,建立了综合的业务上下文信息处理的实体架构和交互式业务提供流程。
(2)其次,在建立了健全的业务上下文信息处理架构基础上,进一步研究基于未来上下文信息的各种预测理论及其关键技术。目前上下文预测主要是针对位置等具体领域的预测,没有建立普遍通用的预测模型且预测准确率不高,针对以上问题,本文提出了基于信任网络和协作过滤算法的业务上下文预测方法,将用户相似度和信任度结合起来,并建立了具有普遍通用的用户-项目-上下文UIC(User-Item-Context)三维协作过滤模型,结合用户的上下文信息进行推理预测。基于业务上下文预测技术解决业务的前摄性问题,为用户提供引导型／推荐型的个性化消费业务奠定基础。
(3)再次,无论是基于当前和历史上下文的上下文感知系统,还是基于未来上下文的上下文预测,上下文信息缺失都是不可避免的难题。本文仅对于传感器感知具体领域,分析了上下文信息这一“流数据”形式特点,充分利用各传感器采集数据之间的关联性,并且结合传感器数据的时空关系特性,提出了基于时空关系和关联规则挖掘的上下文信息缺失插补方法(STARM),全面综合讨论了数据插补方法,提高了传感器数据缺失插补的准确性,并通过温度传感器采集数据验证了这一算法可用性和高效性。
(4)最后,在以上提出的模型架构和相关的理论、方法与算法基础上,本文设计实现了业务上下文处理平台,并初步建立了集成模拟、仿真和实验为一体的业务上下文处理试验床,通过采集模拟传感数据,并进行上下文信息融合处理、推理和预测,演示相应的智能型、前摄型示例业务场景。
The intelligent, personalized service is the development trend of next generation networks. The ability of across a variety of bearer networks and multiple operator domains, universal query access, integration by requirements, context information processing, mobile seamless application will be required and thus develop a user-centric humanization, intelligent service environment. The service context information processing mechanism is prerequisite to provide users with intelligent, personalized services, involving a variety of context information processing, including the homologous or heterogeneous context information processing of network, terminals and user environment. This paper mainly study on the research of service applications level of next generation network, including the future service context information processing mechanism, prediction theory and key technology.
(1) First, for the service context information processing mechanism is not perfect currently, lack of the general and scalable support platform of context information processing. To establish the service context information processing architecture of universal significance, scalable and layered heterogeneous or homogeneous multi-service context information. To establish the scope and level framework of the relevant service context information to support the intelligence service information processing mechanism, and thus to provide users with personalized services. Based on the research of current context-aware technology to establish the entity architecture and interactive service provision model of comprehensive and systematic information.
(2) Secondly, on that basis of the establishment of wholesome service context information processing architecture, further research on various prediction theory and key technologies based on future context.Currently,the context prediction mainly aims at the specific areas for example location, it does not establish the general prediction model and the prediction accuracy is not high, in order to solve the above problem, this paper proposes the approach of context prediction based on trust network and collaborative filtering algorithms, which combines users'trust value into the users'similarity, and establishes the universal general three-dimensional collaborative filtering model of user-item-context, combines with the users'context information to conduct prediction and reasoning. To solve the problems of proactive service based on context prediction technology, to lay the foundation for user service of guiding type or recommendation type personalized.
(3) Again, whether is the context-aware systems based on current and historical context, or the prediction based on future context, the context data missing is an inevitable problem. This paper only aims at the specific areas of sensor-aware, to analyze the flow data form of context information, to make full use of data relevance between every collecting sensor and take spatial-temporal relationship into account, and then proposes an imputation technique for context data missing based on Spatial-Temporal and Association Rule Mining (STARM), to discuss comprehensively the data imputation approach of missing data, to improve the accuracy of imputation of missing data. Finally, the simulation experiment verifies the rationality and efficiency of STARM through temperature sensor data acquisition.
(4) Finally, based on the proposed model framework, related theories, methods and algorithms, this paper designs and implements the service context processing platform, to achieve preliminary the test bed of service context processing with an integrated imitation, simulation and experiment, through the acquisition of analog sensor data and carry out fusion, reasoning and prediction of context information, to demonstrate the service context sample scenarios with proactive, intelligent service.
(1)首先,针对目前业务上下文信息处理机制还不完善,缺乏可扩展的通用的上下文信息处理支撑平台,建立了具有普遍意义、可扩展和分层的多源异质或同质的业务上下文信息的处理机制架构SCIPMA(Service Context Information Processing Mechanism Architecture),确立了业务相关的上下文信息的范畴和层次框架,以支持业务的智能化处理机制,为用户提供个性化业务。在此基础上,建立了综合的业务上下文信息处理的实体架构和交互式业务提供流程。
(2)其次,在建立了健全的业务上下文信息处理架构基础上,进一步研究基于未来上下文信息的各种预测理论及其关键技术。目前上下文预测主要是针对位置等具体领域的预测,没有建立普遍通用的预测模型且预测准确率不高,针对以上问题,本文提出了基于信任网络和协作过滤算法的业务上下文预测方法,将用户相似度和信任度结合起来,并建立了具有普遍通用的用户-项目-上下文UIC(User-Item-Context)三维协作过滤模型,结合用户的上下文信息进行推理预测。基于业务上下文预测技术解决业务的前摄性问题,为用户提供引导型／推荐型的个性化消费业务奠定基础。
(3)再次,无论是基于当前和历史上下文的上下文感知系统,还是基于未来上下文的上下文预测,上下文信息缺失都是不可避免的难题。本文仅对于传感器感知具体领域,分析了上下文信息这一“流数据”形式特点,充分利用各传感器采集数据之间的关联性,并且结合传感器数据的时空关系特性,提出了基于时空关系和关联规则挖掘的上下文信息缺失插补方法(STARM),全面综合讨论了数据插补方法,提高了传感器数据缺失插补的准确性,并通过温度传感器采集数据验证了这一算法可用性和高效性。
(4)最后,在以上提出的模型架构和相关的理论、方法与算法基础上,本文设计实现了业务上下文处理平台,并初步建立了集成模拟、仿真和实验为一体的业务上下文处理试验床,通过采集模拟传感数据,并进行上下文信息融合处理、推理和预测,演示相应的智能型、前摄型示例业务场景。
The intelligent, personalized service is the development trend of next generation networks. The ability of across a variety of bearer networks and multiple operator domains, universal query access, integration by requirements, context information processing, mobile seamless application will be required and thus develop a user-centric humanization, intelligent service environment. The service context information processing mechanism is prerequisite to provide users with intelligent, personalized services, involving a variety of context information processing, including the homologous or heterogeneous context information processing of network, terminals and user environment. This paper mainly study on the research of service applications level of next generation network, including the future service context information processing mechanism, prediction theory and key technology.
(1) First, for the service context information processing mechanism is not perfect currently, lack of the general and scalable support platform of context information processing. To establish the service context information processing architecture of universal significance, scalable and layered heterogeneous or homogeneous multi-service context information. To establish the scope and level framework of the relevant service context information to support the intelligence service information processing mechanism, and thus to provide users with personalized services. Based on the research of current context-aware technology to establish the entity architecture and interactive service provision model of comprehensive and systematic information.
(2) Secondly, on that basis of the establishment of wholesome service context information processing architecture, further research on various prediction theory and key technologies based on future context.Currently,the context prediction mainly aims at the specific areas for example location, it does not establish the general prediction model and the prediction accuracy is not high, in order to solve the above problem, this paper proposes the approach of context prediction based on trust network and collaborative filtering algorithms, which combines users'trust value into the users'similarity, and establishes the universal general three-dimensional collaborative filtering model of user-item-context, combines with the users'context information to conduct prediction and reasoning. To solve the problems of proactive service based on context prediction technology, to lay the foundation for user service of guiding type or recommendation type personalized.
(3) Again, whether is the context-aware systems based on current and historical context, or the prediction based on future context, the context data missing is an inevitable problem. This paper only aims at the specific areas of sensor-aware, to analyze the flow data form of context information, to make full use of data relevance between every collecting sensor and take spatial-temporal relationship into account, and then proposes an imputation technique for context data missing based on Spatial-Temporal and Association Rule Mining (STARM), to discuss comprehensively the data imputation approach of missing data, to improve the accuracy of imputation of missing data. Finally, the simulation experiment verifies the rationality and efficiency of STARM through temperature sensor data acquisition.
(4) Finally, based on the proposed model framework, related theories, methods and algorithms, this paper designs and implements the service context processing platform, to achieve preliminary the test bed of service context processing with an integrated imitation, simulation and experiment, through the acquisition of analog sensor data and carry out fusion, reasoning and prediction of context information, to demonstrate the service context sample scenarios with proactive, intelligent service.
引文
1)终端向业务提供商发出业务请求。
2)业务提供商提供鉴权认证,认证通过向终端发业务验证响应,允许业务访问。
3)业务提供商向应用服务器请求业务逻辑。
4)应用服务器查询上下文服务器上业务上下文服务,获得业务上下文信息。通常有两种方式,一是采取上下文服务定制方式；另一种是调用方式,在业务触发时调用。
5)终端、运营商、用户向上下文服务器提供上下文信息,通过SCIPMA系统,将高层上下文封装成上下文服务存储在目录服务器上,提供应用层应用。
6)应用服务器综合上下文信息运行业务逻辑,为终端提供基于上下文感知的业务。
本文提出了新一代网络环境中通用、可扩展业务上下文信息处理分层机制架构SCIPMA,设计实现了业务上下文信息处理实验平台功能模块和交互式业务提供流程。目前,针对新一代网络环境中业务上下文这方面的研究还不完善,还有很多内容亟待进一步研究,例如上下文的推理机制,特别是对于不确定性的推理,如何建立业务上下文模糊本体,利用上下文模糊本体进行不确定性推理还需要进一步探讨。
[3-1]WEISER M. The computer for the twenty-first century [J]. Scientific American,1991,1(3):265.
[3-2]R. Want, A. Hopper, V. Falcao and J. Gibbons "The Active Badge Location System," ACM Transactions on Information Systems, vol.10, no.1, pp. 91-102,1992.
[3-3]Brumitt, Krumm, Meyers and Shafer "Ubiquitous computing and the role of geometry," Personal Communications, IEEE, vol.7, no.5 SN-1070-9916, pp.41-43,2000.
[3-4]A.K. Dey "Providing Architectural Support for Building Context-A ware Applications," PhD thesis, Georgia Institute of Technology,2001.
[3-5]Gaia Group "http://gaia.cs.uiuc.edu/",2006.
[3-6]H.Chen,"An Intelligent Broker Architecture for Pervasive Context-Aware Systems," PhD thesis, University of Maryland,2004.
[3-7]杨润,孟旭东,糜正琨.基于SIP的上下文感知业务装置,重庆邮电学院学报(自然科学版),2006年4月,第18期第2卷.
[3-8]孙克勇,孟旭东,上下文感知业务在移动网络中的应用,西安邮电学院学报,2006年9月,第11期第5卷.
[3-9]邹文科,孟祥武,基于语义Web技术的上下文感知的智能移动服务,计算机科学,2006年第33卷第9期.
[3-10]http://www.w3.org/Mobile/CCPP/.
[3-11]T. Strang and C. Linnhoff-Popien "A Context Modeling Survey," In First International Workshop on Ad-vanced Context Modeling, Reasoning And Management,UbiComp 2004.
[3-12]N.Ryan "ConteXtML:Exchanging Contextual Information between a Mobile Client and the FieldNote Server," http://www.cs.kent.ac.uk/projects/mobicomp/fnc/ConteXtML.html,2000.
[3-13]K.Henricksen "Developing context-aware pervasive computing applications: Models and approach," Journal of Pervasive and Mobile Computing,2006.
[3-14]A. Schmidt and K.van Laerhoven "How to build smart appliances, "Personal Communications, IEEE, vol.8JA-Personal Com, pp.66EP-71,2001.
[3-15]J.McCarthy "Notes on Formalizing Contexts," Proceedings of the Thirteenth International Joint Conference on Artificial Intelligence, Pp.555-560,1993.
[3-16]Chen et al. SOUPA:Standard Ontology for Ubiquitous and Pervasive Applications, The First Annual International Conference on Mobile and Ubiquitous Systems Networking and Services 2004.
[3-17]http://www.aktors.org/technologies/3store/.
[3-18]ZADEH L.A.Fuzzy sets [J].Information and Control,1965,8(3):338-353.
[3-19]Kuster U,Konig-Ries B,Stern M,et al.DIANE-An Integrated Approach to Automated Service Discovery, Matchmaking and Composition [A].Accepted for publication in:Proceedings of the 16th International World Wide Web Conference (WWW2007) [C], May 2007.
[3-20]Software Building Blocks. Mobile Agent Technologies [R].2005.
[3-21]Bhagwat P.Bluetooth.Technology for Short Range Wireless Apps [J].IEEE Internet Computing. Volume 5, No.3, May/June 2001.
[3-22]Sun Microsystems.Jini Architecture Specification [EB/OL], Revision 1.0.October 2000, available at http://www.sun.com/jini.
[3-23]UPnP Forum.UPnP Device Architecture 1.0[EB/OL], v1.0.1.December2003, available at http://www.upnp.org/resources/documents/ CleanUPnPDA101-20031202s.pdf.
[3-24]Miller B, Salutation R P. Service Discovery in PervasiveComputing Environments[R]. Technical Report, IBM White Paper, February 2000.
[3-25]Maximilien E M, Singh M P. Multiagent System for Dynamic Web Services Selection [A]. Workshop on Service-Oriented Computing and Agent-Based Engineering (SOCABE 2005) [C].July 25-29,2005, Utrecht, The Netherlands.
[3-26]Casati F, Castellanos M, Dayal U, et al. Probabilistic, Context-Sensitive and Goal-Oriented Service Selection [A].Proc.2nd International Conference on Service Oriented Computing (ICSOC04)[C].New York, November 2004.
[3-27]Kalepu S, Krishnaswamy S, and Verity S L:A QoS Metric for Selecting Web Services and Providers [A].Proc.4th Intl.Conf.Web Information Systems Eng.Workshops[C], pp.131-139,2003.
[3-28]Sreenath R M, Singh M P:Agent-based service selection [J]. Journal of Web Semantics (3),2004:261-279.
[3-29]林涛,普适计算中的服务选择问题[D],东北大学博士学位论文,2004.
[3-30]黄润才,庄怡雯,周集良,任务计算中基于语义Web的上下文感知服务选择[J].计算机应用,2009,29(3)：892-895.31.
[3-31]Wang X, Vitvar T, Kerrigan M, et al.A QoS-Aware Selection Model for Semantic Web Services [A]. Service-Oriented Computing[C], ICSOC 2006 (2006), pp.390-401.
[3-32]Tsesmetzis D, Roussaki I, Sykas E.QoS-aware service evaluation and selection [J].European Journal of Operational Research 191(2008), 1101-1112.
[3-33]Ma Y, Zhang C.Quick convergence of genetic algorithm for QoS-driven web service selection [J]. Computer Networks,52 (2008),1093-1104.
[3-34]http://www.informatik.uni-trier.de/～ley/db/indices/a-tree/m/ Maximilien:E=Michael.html.
[3-35]http://www.csc.ncsu.edu/faculty/mpsingh/.
[3-36]http://www.cmpe.boun.edu.tr/～pyolum/.
[3-37]张磊,糜正琨,成际镇,基于Parlay的状态感知通信业务技术及实现,1001-9081(2006)07-1546-03.
从图中可以看出,TCF的MAE值不仅在数据稀疏时偏高,而且在数据较稠密时也无法达到理想的值。这其中主要的原因是传统的协作过滤技术没有考虑到环境中上下文的相似度。而CCF在数据稀疏的情况下,MAE值很高,由于加入了对上下文的考虑,进一步稀释了用户-评价矩阵,使得算法在启动阶段很难找到类似的用户,因此造成了在数据稀疏条件下算法准确性的下降。TNCF在启动时和TCF以及CCF相比具有较好的性能,由此可见,该算法有效的改进了在数据稀疏时预测的准确性。
本文提出了基于信任网络和协作过滤的上下文预测算法,将用户相似度和信任度结合起来,并引入上下文相似度构成用户-项目-上下文三维协作过滤模型来进行上下文预测,并详细描述了该方法的基本框架和预测基本流程。最后通过一系列仿真实验验证了此方法预测的准确率远远胜过传统协作过滤算法。
另外,用户上下文信息还包括用户的社会上下文,特别是用户本身的社会属性,例如年龄,性别,职业等,也是将来用户预测推荐系统中我们需要考虑的重要因素,此外,本文考虑用户的信任度取值区间仅限于[0,1]之内,在[-1,1]区间内上下文预测的研究还需要进一步探讨。
[4-1]Dey A.K., The Context Toolkit a Toolkit for Context-aware Applications,http://www.cs.berkeley.edu/-dey/context.html,2001.
[4-2]Chen H., Finin T., Joshi A.An ontology for context-aware pervasive computing environments. The Knowledge Engineering Review, Cambridge University Press,2003, (18),197-207.
[4-3]Fahy P., Clarke S., CASS:a middleware for mobile context-aware applications, Workshop on Context Awareness, MobiSys,2004.
[4-4]Gu T.,Pung H.K.,Zhang D.Q.,A middleware for building context-aware mobile services, In proceedings of IEEE Vehicular Technology Conference(VTC),Milan,Italy,2004.
[4-5]Mayrhofer R.,Radi H.,Ferscha A., Recognizing and predicting context by learning from user behavior, In Proceedings of the International Conference on Advances in Mobile Multimedia(MoMM),Austrian Computer Society(OCG),2003,22-35.
[4-6]Mayrhofer R.,An architecture for context prediction, In Advances in Pervasive Computing, Austrian Computer Society(OCG),2004,(176),65-72.
[4-7]Mayrhofer R., an Architecture for Context Prediction, [PhD Dissertation], University of Linz,2004.
[4-8]Laasonen K., Raento M., Toivonen H., Adaptive on-device location recognition, In Proceedings of the Second International Conference on Pervasive Computing, Springer-Verlag,2004,287-304.
[4-9]Patterson D.J., Lin L., Fox D., etc.Inferring High-Level Behavior from Low-Level Sensors.In 5th International Conference on Ubiquitous Computing, Seattle, WA, USA,2003,73-89.
[4-10]Kaowthumrong K., Lebsack J., Han R., Automated selection of the active device in Interactive multi-device smart spaces, In Proceedings of the Ubicomp'02 Workshop on Supporting Spontaneous Interaction in Ubiquitous Computing Settings,2002.
[4-11]Petzold J.,Bagci F.,Trumler W.,Ungerer T.,Vintan,L,Global state context prediction techniques applied to a smart office building, In Proceedings of the Communication Networks and Distributed Systems Modeling and Simulation Conference(CNDS),2004.
[4-12]S. Sigg, S. Haseloff, and K. David. A novel approach to context prediction in ubicomp environments. The 17th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC'06).
[4-13]H. Si, Y. Kawahara, H. Kurasawa, H. Morikawa, and T. Aoyama. A context-aware collaborative filtering algorithm for real world oriented content delivery service. In Ubicom2005 Metapolis and Urban Life Workshop,2005.
[4-14]Zhdanova, A.V., Davies, M., Jorns, O., Stavroulaki, V., Demestichas, P., Gonzalez, M., Moessner, K., Carrez, F., Rajasekaran, H., Lo Iacono, L. "MOSS:Mobile Social Spaces". In Proceedings of the 1st International Workshop on Blending Physical and Digital Spaces on the Internet (OneSpace2008) at the 1st Future Internet Symposium (FIS'08),28-30 September 2008, Vienna, Austria (2008).
[4-15]A. Chen, "Context-Aware Collaborative Filtering System", Proc. of the International Workshop on Location-and Context-Awareness 2005, Springer LNCS 3479, Berlin/Heidelberg,2005, pp.244-253.
[4-16]W. Woerndl, C. Schueller, and R. Wojtech, "A Hybrid Recommender System for Context-aware Recommendations of Mobile Applications", IEEE 3rd International Workshop on Web Personalization, Recommender Systems and Intelligent User Interfaces (WPRSIUI'07), Istanbul, Turkey, Apr.2007.
[4-17]Adomavicius, G, Sankaranarayanan, R., Sen, S., and Tuzhilin, A.2005. Incorporating contextual information in recommender systems using a multidimensional approach. ACM Trans. Inf. Syst.23,1 (Jan.2005), 103-145.
[4-18]卢竹兵,唐雁,一种基于信任网络的协同过滤推荐策略,西南师范大学学报(自然科学版),Vol(33),No(2),2008年4月.
[4-19]Michael C.Mozer., The Neural Network House:An Environment that Adapts to its Inhabitants, In AAAI Spring Symposium on Intelligent Environments, Menlo Park CA USA,1998,110-114.
[4-20]Ashbrook D., Starner T., Using GPS to learn significant locations and predict movement across multiple users, Personal and Ubiquitous Computing,2003 7(5),275-286.
[4-21]K.Cheverst, N.Davies, K.Mitchell, A. Friday, and C.Efstratiou, Developing a Context-aware Electronic Tourist Guide:Some Issues and Experiences, In Proceedings of CHI2000, Netherlands,2000,17-24.
[4-22]Petteri N,Miquel M,John A F.Enabling proactiveness through context prediction[C]//Proceedings of the Workshop on Context Awareness for Proactive Systems(CAPS 2005).Helsinki,Finland,2005.
从图5-3可以看出,EM和SLR由于算法的简单性所占用的内存最小,而与其它算法比较中STARM效率较高,占用内存较小,从而节省了系统空间开销。
本文依据传感器数据这一“流数据”的关联性和时空特性,引入了时间新鲜度的概念,并依据传统的关联规则挖掘的基本过程,提出了基于时空关系和关联规则挖掘的上下文信息缺失插补方法,较传统的缺失数据插补方法具有更高的准确性,并且减小了时空开销。实验证明此方法在数据插补准确性优于传统的统计方法和其它关联规则挖掘方法,验证了算法的合理性和高效性。
由于对传感器数据进行时间序列化,如果采用指数加权,其数值会呈指数级递增,数据有可能会溢出,如何进行控制数据的溢出,是本文未来进一步的研究工作。
[5-1]Cool A L. A review of methods for dealing with missing data[C]. Paper presented at the Annual Meeting of the Southwest Educational Research Association, Dallas, TX,2000,1-34.
[5-2]Shao Jun, Wang Han-sheng.Confidence intervals based on survey data with nearest neighbor imputation [J]. Statistica Sinica, Vol.18, pp.281-297,2008.
[5-3]Gu Dong-bing. Distributed EM algorithm for gaussian mixtures in sensor networks [J], IEEE transactions on neural networks,2008,19(7):1154-1166.
[5-4]Allison, P.D.:Missing data [D]. Thousand Oaks, CA Sage (2002).
[5-5]Qin Yong-song, Zhang Shi-chao.Empirical likelihood confidence intervals for differences between two datasets with missing data [J].Pattern Recognition Letters,2008,29(6):803-812.
[5-6]庞新生,分层随机抽样条件下缺失数据的多重插补方法[J].统计与信息论坛,2009 24(5),19-21.
[5-7]金勇进,邵军.缺失数据的统计处理.北京：中国统计出版社,2009：155-161.
[5-8]Deshpande A, Guestrin C, Madden S, Hellerstein J, Hong W.Model-driven data acquisition in sensor networks [C]. Proceedings of the 30th VLDB Conference, Toronto, Canada,2004,588-599.
[5-9]Le Gruenwald, Hamed Chok, Mazen Aboukhamis. Using data mining to estimate missing sensor data[C]. Proceedings of the Seventh IEEE International Conference on Data Mining Workshops, Norman U.S.A,2007, 207-212.
[5-10]Nan Jiang.A data imputation model in sensor databases [C]. High Performance Computing and Communications, Third International Conference, HPCC 2007, Houston, USA, September 26-28,2007,86-96.
[5-11]Li Y and Parker L E. Classification with missing data in a wireless sensor network[C].IEEE Southeast Conference, Huntsville, Alabama, April 2008, 533-538.
[5-12]Li Y and Parker L E. A spatial-temporal imputation technique for classification with missing data in a wireless sensor network[C].2008 IEEE/RSJ International Conference on Intelligent Robots and Systems, Acropolis Convention Center Nice, France, Sept,22-26,2008,3272-3279.
[5-13]Rubin, R.J.A.Inerence and missing data.Biometrika,63:581-92,1976.
2)业务提供商提供鉴权认证,认证通过向终端发业务验证响应,允许业务访问。
3)业务提供商向应用服务器请求业务逻辑。
4)应用服务器查询上下文服务器上业务上下文服务,获得业务上下文信息。通常有两种方式,一是采取上下文服务定制方式；另一种是调用方式,在业务触发时调用。
5)终端、运营商、用户向上下文服务器提供上下文信息,通过SCIPMA系统,将高层上下文封装成上下文服务存储在目录服务器上,提供应用层应用。
6)应用服务器综合上下文信息运行业务逻辑,为终端提供基于上下文感知的业务。
本文提出了新一代网络环境中通用、可扩展业务上下文信息处理分层机制架构SCIPMA,设计实现了业务上下文信息处理实验平台功能模块和交互式业务提供流程。目前,针对新一代网络环境中业务上下文这方面的研究还不完善,还有很多内容亟待进一步研究,例如上下文的推理机制,特别是对于不确定性的推理,如何建立业务上下文模糊本体,利用上下文模糊本体进行不确定性推理还需要进一步探讨。
[3-1]WEISER M. The computer for the twenty-first century [J]. Scientific American,1991,1(3):265.
[3-2]R. Want, A. Hopper, V. Falcao and J. Gibbons "The Active Badge Location System," ACM Transactions on Information Systems, vol.10, no.1, pp. 91-102,1992.
[3-3]Brumitt, Krumm, Meyers and Shafer "Ubiquitous computing and the role of geometry," Personal Communications, IEEE, vol.7, no.5 SN-1070-9916, pp.41-43,2000.
[3-4]A.K. Dey "Providing Architectural Support for Building Context-A ware Applications," PhD thesis, Georgia Institute of Technology,2001.
[3-5]Gaia Group "http://gaia.cs.uiuc.edu/",2006.
[3-6]H.Chen,"An Intelligent Broker Architecture for Pervasive Context-Aware Systems," PhD thesis, University of Maryland,2004.
[3-7]杨润,孟旭东,糜正琨.基于SIP的上下文感知业务装置,重庆邮电学院学报(自然科学版),2006年4月,第18期第2卷.
[3-8]孙克勇,孟旭东,上下文感知业务在移动网络中的应用,西安邮电学院学报,2006年9月,第11期第5卷.
[3-9]邹文科,孟祥武,基于语义Web技术的上下文感知的智能移动服务,计算机科学,2006年第33卷第9期.
[3-10]http://www.w3.org/Mobile/CCPP/.
[3-11]T. Strang and C. Linnhoff-Popien "A Context Modeling Survey," In First International Workshop on Ad-vanced Context Modeling, Reasoning And Management,UbiComp 2004.
[3-12]N.Ryan "ConteXtML:Exchanging Contextual Information between a Mobile Client and the FieldNote Server," http://www.cs.kent.ac.uk/projects/mobicomp/fnc/ConteXtML.html,2000.
[3-13]K.Henricksen "Developing context-aware pervasive computing applications: Models and approach," Journal of Pervasive and Mobile Computing,2006.
[3-14]A. Schmidt and K.van Laerhoven "How to build smart appliances, "Personal Communications, IEEE, vol.8JA-Personal Com, pp.66EP-71,2001.
[3-15]J.McCarthy "Notes on Formalizing Contexts," Proceedings of the Thirteenth International Joint Conference on Artificial Intelligence, Pp.555-560,1993.
[3-16]Chen et al. SOUPA:Standard Ontology for Ubiquitous and Pervasive Applications, The First Annual International Conference on Mobile and Ubiquitous Systems Networking and Services 2004.
[3-17]http://www.aktors.org/technologies/3store/.
[3-18]ZADEH L.A.Fuzzy sets [J].Information and Control,1965,8(3):338-353.
[3-19]Kuster U,Konig-Ries B,Stern M,et al.DIANE-An Integrated Approach to Automated Service Discovery, Matchmaking and Composition [A].Accepted for publication in:Proceedings of the 16th International World Wide Web Conference (WWW2007) [C], May 2007.
[3-20]Software Building Blocks. Mobile Agent Technologies [R].2005.
[3-21]Bhagwat P.Bluetooth.Technology for Short Range Wireless Apps [J].IEEE Internet Computing. Volume 5, No.3, May/June 2001.
[3-22]Sun Microsystems.Jini Architecture Specification [EB/OL], Revision 1.0.October 2000, available at http://www.sun.com/jini.
[3-23]UPnP Forum.UPnP Device Architecture 1.0[EB/OL], v1.0.1.December2003, available at http://www.upnp.org/resources/documents/ CleanUPnPDA101-20031202s.pdf.
[3-24]Miller B, Salutation R P. Service Discovery in PervasiveComputing Environments[R]. Technical Report, IBM White Paper, February 2000.
[3-25]Maximilien E M, Singh M P. Multiagent System for Dynamic Web Services Selection [A]. Workshop on Service-Oriented Computing and Agent-Based Engineering (SOCABE 2005) [C].July 25-29,2005, Utrecht, The Netherlands.
[3-26]Casati F, Castellanos M, Dayal U, et al. Probabilistic, Context-Sensitive and Goal-Oriented Service Selection [A].Proc.2nd International Conference on Service Oriented Computing (ICSOC04)[C].New York, November 2004.
[3-27]Kalepu S, Krishnaswamy S, and Verity S L:A QoS Metric for Selecting Web Services and Providers [A].Proc.4th Intl.Conf.Web Information Systems Eng.Workshops[C], pp.131-139,2003.
[3-28]Sreenath R M, Singh M P:Agent-based service selection [J]. Journal of Web Semantics (3),2004:261-279.
[3-29]林涛,普适计算中的服务选择问题[D],东北大学博士学位论文,2004.
[3-30]黄润才,庄怡雯,周集良,任务计算中基于语义Web的上下文感知服务选择[J].计算机应用,2009,29(3)：892-895.31.
[3-31]Wang X, Vitvar T, Kerrigan M, et al.A QoS-Aware Selection Model for Semantic Web Services [A]. Service-Oriented Computing[C], ICSOC 2006 (2006), pp.390-401.
[3-32]Tsesmetzis D, Roussaki I, Sykas E.QoS-aware service evaluation and selection [J].European Journal of Operational Research 191(2008), 1101-1112.
[3-33]Ma Y, Zhang C.Quick convergence of genetic algorithm for QoS-driven web service selection [J]. Computer Networks,52 (2008),1093-1104.
[3-34]http://www.informatik.uni-trier.de/～ley/db/indices/a-tree/m/ Maximilien:E=Michael.html.
[3-35]http://www.csc.ncsu.edu/faculty/mpsingh/.
[3-36]http://www.cmpe.boun.edu.tr/～pyolum/.
[3-37]张磊,糜正琨,成际镇,基于Parlay的状态感知通信业务技术及实现,1001-9081(2006)07-1546-03.
从图中可以看出,TCF的MAE值不仅在数据稀疏时偏高,而且在数据较稠密时也无法达到理想的值。这其中主要的原因是传统的协作过滤技术没有考虑到环境中上下文的相似度。而CCF在数据稀疏的情况下,MAE值很高,由于加入了对上下文的考虑,进一步稀释了用户-评价矩阵,使得算法在启动阶段很难找到类似的用户,因此造成了在数据稀疏条件下算法准确性的下降。TNCF在启动时和TCF以及CCF相比具有较好的性能,由此可见,该算法有效的改进了在数据稀疏时预测的准确性。
本文提出了基于信任网络和协作过滤的上下文预测算法,将用户相似度和信任度结合起来,并引入上下文相似度构成用户-项目-上下文三维协作过滤模型来进行上下文预测,并详细描述了该方法的基本框架和预测基本流程。最后通过一系列仿真实验验证了此方法预测的准确率远远胜过传统协作过滤算法。
另外,用户上下文信息还包括用户的社会上下文,特别是用户本身的社会属性,例如年龄,性别,职业等,也是将来用户预测推荐系统中我们需要考虑的重要因素,此外,本文考虑用户的信任度取值区间仅限于[0,1]之内,在[-1,1]区间内上下文预测的研究还需要进一步探讨。
[4-1]Dey A.K., The Context Toolkit a Toolkit for Context-aware Applications,http://www.cs.berkeley.edu/-dey/context.html,2001.
[4-2]Chen H., Finin T., Joshi A.An ontology for context-aware pervasive computing environments. The Knowledge Engineering Review, Cambridge University Press,2003, (18),197-207.
[4-3]Fahy P., Clarke S., CASS:a middleware for mobile context-aware applications, Workshop on Context Awareness, MobiSys,2004.
[4-4]Gu T.,Pung H.K.,Zhang D.Q.,A middleware for building context-aware mobile services, In proceedings of IEEE Vehicular Technology Conference(VTC),Milan,Italy,2004.
[4-5]Mayrhofer R.,Radi H.,Ferscha A., Recognizing and predicting context by learning from user behavior, In Proceedings of the International Conference on Advances in Mobile Multimedia(MoMM),Austrian Computer Society(OCG),2003,22-35.
[4-6]Mayrhofer R.,An architecture for context prediction, In Advances in Pervasive Computing, Austrian Computer Society(OCG),2004,(176),65-72.
[4-7]Mayrhofer R., an Architecture for Context Prediction, [PhD Dissertation], University of Linz,2004.
[4-8]Laasonen K., Raento M., Toivonen H., Adaptive on-device location recognition, In Proceedings of the Second International Conference on Pervasive Computing, Springer-Verlag,2004,287-304.
[4-9]Patterson D.J., Lin L., Fox D., etc.Inferring High-Level Behavior from Low-Level Sensors.In 5th International Conference on Ubiquitous Computing, Seattle, WA, USA,2003,73-89.
[4-10]Kaowthumrong K., Lebsack J., Han R., Automated selection of the active device in Interactive multi-device smart spaces, In Proceedings of the Ubicomp'02 Workshop on Supporting Spontaneous Interaction in Ubiquitous Computing Settings,2002.
[4-11]Petzold J.,Bagci F.,Trumler W.,Ungerer T.,Vintan,L,Global state context prediction techniques applied to a smart office building, In Proceedings of the Communication Networks and Distributed Systems Modeling and Simulation Conference(CNDS),2004.
[4-12]S. Sigg, S. Haseloff, and K. David. A novel approach to context prediction in ubicomp environments. The 17th Annual IEEE International Symposium on Personal, Indoor and Mobile Radio Communications (PIMRC'06).
[4-13]H. Si, Y. Kawahara, H. Kurasawa, H. Morikawa, and T. Aoyama. A context-aware collaborative filtering algorithm for real world oriented content delivery service. In Ubicom2005 Metapolis and Urban Life Workshop,2005.
[4-14]Zhdanova, A.V., Davies, M., Jorns, O., Stavroulaki, V., Demestichas, P., Gonzalez, M., Moessner, K., Carrez, F., Rajasekaran, H., Lo Iacono, L. "MOSS:Mobile Social Spaces". In Proceedings of the 1st International Workshop on Blending Physical and Digital Spaces on the Internet (OneSpace2008) at the 1st Future Internet Symposium (FIS'08),28-30 September 2008, Vienna, Austria (2008).
[4-15]A. Chen, "Context-Aware Collaborative Filtering System", Proc. of the International Workshop on Location-and Context-Awareness 2005, Springer LNCS 3479, Berlin/Heidelberg,2005, pp.244-253.
[4-16]W. Woerndl, C. Schueller, and R. Wojtech, "A Hybrid Recommender System for Context-aware Recommendations of Mobile Applications", IEEE 3rd International Workshop on Web Personalization, Recommender Systems and Intelligent User Interfaces (WPRSIUI'07), Istanbul, Turkey, Apr.2007.
[4-17]Adomavicius, G, Sankaranarayanan, R., Sen, S., and Tuzhilin, A.2005. Incorporating contextual information in recommender systems using a multidimensional approach. ACM Trans. Inf. Syst.23,1 (Jan.2005), 103-145.
[4-18]卢竹兵,唐雁,一种基于信任网络的协同过滤推荐策略,西南师范大学学报(自然科学版),Vol(33),No(2),2008年4月.
[4-19]Michael C.Mozer., The Neural Network House:An Environment that Adapts to its Inhabitants, In AAAI Spring Symposium on Intelligent Environments, Menlo Park CA USA,1998,110-114.
[4-20]Ashbrook D., Starner T., Using GPS to learn significant locations and predict movement across multiple users, Personal and Ubiquitous Computing,2003 7(5),275-286.
[4-21]K.Cheverst, N.Davies, K.Mitchell, A. Friday, and C.Efstratiou, Developing a Context-aware Electronic Tourist Guide:Some Issues and Experiences, In Proceedings of CHI2000, Netherlands,2000,17-24.
[4-22]Petteri N,Miquel M,John A F.Enabling proactiveness through context prediction[C]//Proceedings of the Workshop on Context Awareness for Proactive Systems(CAPS 2005).Helsinki,Finland,2005.
从图5-3可以看出,EM和SLR由于算法的简单性所占用的内存最小,而与其它算法比较中STARM效率较高,占用内存较小,从而节省了系统空间开销。
本文依据传感器数据这一“流数据”的关联性和时空特性,引入了时间新鲜度的概念,并依据传统的关联规则挖掘的基本过程,提出了基于时空关系和关联规则挖掘的上下文信息缺失插补方法,较传统的缺失数据插补方法具有更高的准确性,并且减小了时空开销。实验证明此方法在数据插补准确性优于传统的统计方法和其它关联规则挖掘方法,验证了算法的合理性和高效性。
由于对传感器数据进行时间序列化,如果采用指数加权,其数值会呈指数级递增,数据有可能会溢出,如何进行控制数据的溢出,是本文未来进一步的研究工作。
[5-1]Cool A L. A review of methods for dealing with missing data[C]. Paper presented at the Annual Meeting of the Southwest Educational Research Association, Dallas, TX,2000,1-34.
[5-2]Shao Jun, Wang Han-sheng.Confidence intervals based on survey data with nearest neighbor imputation [J]. Statistica Sinica, Vol.18, pp.281-297,2008.
[5-3]Gu Dong-bing. Distributed EM algorithm for gaussian mixtures in sensor networks [J], IEEE transactions on neural networks,2008,19(7):1154-1166.
[5-4]Allison, P.D.:Missing data [D]. Thousand Oaks, CA Sage (2002).
[5-5]Qin Yong-song, Zhang Shi-chao.Empirical likelihood confidence intervals for differences between two datasets with missing data [J].Pattern Recognition Letters,2008,29(6):803-812.
[5-6]庞新生,分层随机抽样条件下缺失数据的多重插补方法[J].统计与信息论坛,2009 24(5),19-21.
[5-7]金勇进,邵军.缺失数据的统计处理.北京：中国统计出版社,2009：155-161.
[5-8]Deshpande A, Guestrin C, Madden S, Hellerstein J, Hong W.Model-driven data acquisition in sensor networks [C]. Proceedings of the 30th VLDB Conference, Toronto, Canada,2004,588-599.
[5-9]Le Gruenwald, Hamed Chok, Mazen Aboukhamis. Using data mining to estimate missing sensor data[C]. Proceedings of the Seventh IEEE International Conference on Data Mining Workshops, Norman U.S.A,2007, 207-212.
[5-10]Nan Jiang.A data imputation model in sensor databases [C]. High Performance Computing and Communications, Third International Conference, HPCC 2007, Houston, USA, September 26-28,2007,86-96.
[5-11]Li Y and Parker L E. Classification with missing data in a wireless sensor network[C].IEEE Southeast Conference, Huntsville, Alabama, April 2008, 533-538.
[5-12]Li Y and Parker L E. A spatial-temporal imputation technique for classification with missing data in a wireless sensor network[C].2008 IEEE/RSJ International Conference on Intelligent Robots and Systems, Acropolis Convention Center Nice, France, Sept,22-26,2008,3272-3279.
[5-13]Rubin, R.J.A.Inerence and missing data.Biometrika,63:581-92,1976.