网络环境下高校图书馆智能咨询服务模式研究

Research on the Intelligent Consulting Service Model of University Library in the Network Environment

高校图书馆智能发展过程中,依靠简单的数字咨询模式实现咨询,只能提供被动静态服务,使得咨询服务结果的命中率(HR)较低。因此,在网络环境下提出一种新的高校图书馆智能咨询服务模式。抓取大量高校图书馆咨询记录数据后,运用知识图谱概念对其进行体系化处理,建立智能咨询服务领域知识库。根据用户行为数据和咨询记录,主动感知用户咨询需求并将其准确表达出来。以引入注意力机制的卷积神经网络(CNN)为基础,构建一种智慧问答服务体系,对咨询内容进行深度学习从知识库中找到对应的答案。最后,结合用户画像、图书馆资源画像和情境画像,为咨询用户提供个性化智能推送服务。应用分析结果表明:所研究的咨询服务模式给出的咨询服务结果命中率超过了80%,非常好地满足了图书馆智能发展要求。

In the process of intelligent development of university libraries,relying on a simple digital consultation model only enables passive static services,resulting in a lower hit rate(HR)of consultation service outcomes.Therefore,a new intelligent consultation service model for university libraries is proposed in the network environment.By aggregating a large amount of consultation records data from university libraries and utilizing the concepts of knowledge graph,a systematic treatment is applied to establish a knowledge base in the field of intelligent consultation services.Based on user behavior data and consultation records,the user's consultation needs are proactively sensed and accurately expressed.A smart question-answering service system is built on the foundation of Convolutional Neural Network(CNN)with attention mechanisms,to deeply learn and find corresponding answers from the knowledge base.Finally,by combining user profile,library resource profile,and contextual profile,personalized intelligent push services are provided to consulting users.Application analysis results demonstrate that the researched consultation service model achieves a hit rate of over 80%,effectively meeting the requirements of intelligent development in university libraries.