项目名称: 信息网络市场中面向主题域识别的博弈融合方法
项目编号: No.61472070
项目类型: 面上项目
立项/批准年度: 2015
项目学科: 自动化技术、计算机技术
项目作者: 寇月
作者单位: 东北大学
项目金额: 82万元
中文摘要: 资源空间和用户空间中的所有信息网络构成了信息网络市场,它能够形象地体现供需双方匹配的条件与互动。基于信息网络市场的主题域识别存在许多新的特点,而传统技术仅基于单一信息网络,对信息网络动态性的适应能力相对薄弱,在匹配中忽略了供需双方的互动。为此,本课题将针对信息网络市场中面向主题域识别的博弈融合方法进行研究,主要包括:研究面向融合的资源空间构建模型,将实体主题特征建模为知识,支持知识智能化挖掘、推演、优化及增量式维护;研究面向融合的用户空间构建模型,基于用户静态特征和动态特征感知用户兴趣;研究基于博弈论的信息网络多层面融合模型,以知识质量和用户利益为驱动实现知识博弈融合和兴趣博弈融合;研究基于博弈融合的主题域匹配策略,支持供需匹配市场中多种主题域匹配方式。本课题的研发成果将为管理开放的数据资源、开发新型的实体搜索理论提供良好的支持,达到提高资源利用率和工作效率的目的,具有广阔的前景。
中文关键词: 主题域识别;信息网络市场;博弈融合
英文摘要: All the information networks in resource space and user space make up an information network market, which can reflect the matching and interaction between supply and demand. There are some new features during topic domain identification based on the information network market. However, traditional identification techniques are only based on a single information network, which is deficinet in analyzing the entities and the relationships more fully. Their ability to adapt to the dynamic updates of information network is lower, also the interaction between supply and demand is neglected during matching. Therefore, in this project we will research the game fusion approach for topic domain identification in information network market. The research content includes: A model for resource space construction oriented to fusion will be presented, which can support intelligent knowledge mining, deriving, optimizing and maintaining incrementally. A model for user space construction oriented to fusion will be presented, which can identity user's interest based on his static and dynamic characteristics. A game theory based fusion model for multiple information networks will be proposed. Drived by knowledge quality and user benefit, it can implement knowledge game fusion and interest game fusion. Some strategies for topic domain matching based on game fusion will be researched, which can support multiple matching ways in the market between supply and demand. The result of the project will provide better support for managing open resources and developing new theory about entity search, also will facilitate higher usage rate of resources.
英文关键词: Topic Domain Identification;Information Network Market;Game Fusion