中文摘要：

[目的／意义]大数据背景下，如何构建合适的用户行为模型并基于海量的行为日志数据提供个性化服务，是当前图书馆大数据应用落地迫切需要解决的问题。[方法/过程]首先分析用户行为模型构建的研究现状及存在的困境，接着密切结合大数据背景下个性化服务的特征，针对性提出基于本体的高校图书馆用户行为模型的构建策略和构建方法，并设计一种利用用户日志库提取用户显性兴趣和隐性需求本体的个性化服务方案，最后给出基于流行的Hadoop大数据分析平台和MapReduce计算框架的图书馆个性化服务的应用案例。[结果／结论]基于本体构建的用户行为模型，技术上可与大数据分析平台实现无缝对接，从而提供实时而精准的服务，能有效应对当前大数据环境下图书馆个性化服务面临的“知识迷航”“信息过载”和“情感缺失”的挑战。

英文摘要：

[ Purpose/significance ] It is an urgent problem for libraries big data application to model user behaviors and provide personalized services based on amass data of user logs under big data background. [ Method/process] Firstly, the research status and current dilemma of user behavior model is analyzed, then considering the characteristic of personalized services under big data background, the modeling strategy and method for library user behavior are put forward based on ontology technology, and a personalized service scheme is proposed after the ontology words of user explicit interests and implicit requirements have been extracted from user logs database, and finally an application case for library personalized services is provided based on Hadoop platform and MapReduce framework. [ Result/conclusion] The proposed user behavior ontology model can be seamlessly combined with analysis platform of big data, providing personalized services timely and precisely, and responding the challenges of current library personalized services such as ＂ knowledge loss＂,＂ Information overload＂ and ＂ emotional deficit＂ under big data background.