基于概念图的信息检索的查询扩展模型

针对传统的基于关键词匹配的信息检索存在的查全率和精确率不高的问题,提出一种基于概念图匹配的查询扩展方法：一方面通过知网对用户查询的词或者句子进行扩展后,将用户查询和文档生成概念图;另一方面利用概念图的不完全匹配和语义相似度的计算方法计算概念图的相似度,以提高检索效果。实验结果表明该方法取得了良好的效果。     

基于日志挖掘的查询概念图扩展

提出将概念图引入查询扩展,从概念的层面上进行语义的扩展。使用概念图表示查询可以更准确地表明用户的查询意图,并在此基础上进行语义的扩展,通过这种方法给出的扩展查询更符合用户的查询意图。对用户查询进行基于概念图的查询扩展,并将结果与百度的相似查询进行了比较,证明基于概念图的查询扩展能更准确地把握用户的查询意图。     

基于概念图匹配的语义相似性算法研究

本文针对传统搜索技术查全率和查准率不能满足用户日益增长的需求这一突出问题,提出一种基于概念图语义匹配的方法来计算两个本体中类之间的相似性,文中提到的本体是由实体类、这些类之间的语义关系和描述这些类的不同特征组成的.该模型首先将用户的查询信息转变为一个概念图,然后和已有的资源概念图进行匹配计算语义的相似性,实例表明该方法可以满足用户的需求,提高了检索效率.     

基于领域本体的贝叶斯网络检索模型研究

针对传统的信息检索方法无法实现用户查询的语义理解、检索效率低等问题,本文提出基于领域本体进行查询扩展的贝叶斯网络检索模型。该模型首先将用户查询通过领域本体进行语义扩展,然后将扩展后的查询作为证据在贝叶斯网络检索模型中进行传播,进而得到查询结果,实验表明本文提出的贝叶斯网络检索模型能提高检索效率。     

支持用户偏好查询的领域概念图模型

针对目前的领域概念查询聚类方法中未见考虑用户偏好,提出一种支持用户偏好查询的领域概念图模型.该图模型主要包括两部分:基于概念本身考虑,利用综合语义相似度计算方法构建概念的语义关系图;基于用户查询偏好考虑,采用改进的互信息计算用户生成数据间隐含的查询偏好,将其结果用于补全领域概念的语义关系图.这一处理过程使得原有领域概念...     

教学资源概念查询扩展研究

查询扩展是在原查询词的基础上加入与用户查询词相关的词或者词组,组成新的、更准确的查询序列,使扩展后的查询序列能更清晰地表达用户的查询请求,克服自然语言的“二义性”。基于《计算机网络》概念语义网络能更加有效地找出计算机网络领域内查询词的概念词及扩展概念词,并向上拓展将各个查询词的原始语义关系联接起来,解决了查询词之间缺乏联系的问题,为扩展检索的实现奠定基础。描述了概念语义网络的生成方法、关联概念树的抽取方法和查询扩展检索的计算机实现流程,为教学资源领域的在线学习提供了技术支持。     

语义桌面环境下一种基于概念图的文档检索方法

语义桌面作为语义Web的一个重要分支,可以为个人计算机用户提供丰富的元数据,用以记录桌面文档的各种特征。这些特征包括文件的常规属性和与用户行为相关的属性,它们为桌面文档检索提供帮助。受到"概念图"理论的启发,本文提出了一种语义桌面环境下的文档检索算法。该算法能有效地利用语义桌面提供的元数据建立一种便于快速查找的文档索引结构,迅速地确定用户查询与桌面文档之间的投影算子。实验表明,该算法的时间效率比以往的基于匹配推理的投影算法有很大的提高,可以在很大程度上满足用户对桌面文档进行快速检索的需求。     

基于文本与语义相关性分析的图像检索

为了更加有效地检索到符合用户复杂语义需求的图像，提出一种基于文本描述与语义相关性分析的图像检索算法。该方法将图像检索分为两步：基于文本语义相关性分析的图像检索和基于SIFT特征的相似图像扩展检索。根据自然语言处理技术分析得到用户文本需求中的关键词及其语义关联，在选定图像库中通过语义相关性分析得到“种子”图像；接下来在图像扩展检索中，采用基于SIFT特征的相似图像检索，利用之前得到的“种子”图像作为查询条件，在网络图像库中进行扩展检索，并在结果集上根据两次检索的图像相似度进行排序输出，最终得到更加丰富有效的图像检索结果。为了证明算法的有效性，在标准数据集Corel5K和网络数据集Deriantart8K上完成了多组实验，实验结果证明该方法能够得到较为精确地符合用户语义要求的图像检索结果，并且通过扩展算法可以得到更加丰富的检索结果。     

NN型复合结构的语义关系识别及相似度计算

目前信息检索的正确率不太高,原因之一是用现有的检索模型难以表示完整的用户查询意图,而用户在查询中大量使用了复合结构.通过实例探索了汉语NN型复合结构基于概念图的语义关系标引,发现复合结构的关联语义关系可以通过子成分的上下文求解.这些上下文通过网络进行识别抽取,并借助<同义词词林>进行泛化以解决数据稀疏性问题.复合结构内部的语义关系用向量来表示,向量的每一维代表了能表示复合结构语义关系的一个上下文.实验表明,提出的方法取得了较好的结果.     

云计算下支持语义的可搜索加密方法研究

传统的云计算下的可搜索加密算法没有对查询关键词进行语义扩展,导致了用户查询意图与返回结果存在语义偏差,并且对检索结果的相关度排序不够合理,无法满足用户对智能搜索的需求。对此,提出了一种支持语义的可搜索加密方法。该方法利用本体知识库实现了用户查询的语义拓展,并通过语义相似度来控制扩展词的个数,防止因拓展词过多影响检索的精确度。同时,该方法利用文档向量、查询向量分块技术构造出对应的标记向量,以过滤无关文档,并在查询-文档的相似度得分中引入了语义相似度、关键词位置加权评分及关键词-文档相关度等影响因子,实现了检索结果的有效排序。实验结果表明,该方法在提高检索效率的基础上显著改善了检索结果的排序效果,提高了用户满意度。     

基于兴趣模型的查询扩展

针对信息检索中查询与文档集之间可能存在的“词不匹配”问题,基于兴趣模型提出一种将概念化的兴趣知识与向量空间模型相结合的查询扩展方法。该方法能根据阈值来判断查询扩展是否可行。用户的兴趣偏好是通过Agent代理实时获取的,兴趣知识采用HNC（Hierarchical Network of Concepts, 概念层次网络）理论的概念符号体系表达,这样便于计算概念之间的相似度。实验结果表明,经过查询扩展后的结果相对于未加入查询扩展的结果在性能上提高了29.1%。     

Designing for Semantic Access: A Video Browsing System

Users of browsing applications often have vague information needs which can only be described in conceptual terms. Therefore, a video browsing system must accept conceptual queries for preselection and offer mechanisms for interactive inspection of the result set by the user. In this paper, we describe a MM-DBMS that we extended with the following components: Our retrieval engine calculates relevance values for the results of a conceptual query by feature aggregation on video shot granularity to offer conceptual, content-based access. To reduce startup delays within sessions, our admission control module admits only complete browsing sessions, if required resources, which are heuristically predicted from query results, are available. In addition, our intelligent client buffer strategy employs the retrieval relevance values to enable flexible user interactions during browsing.     

Deep Learning and Semantic Concept Spaceare Used in Query Expansion

In the practice of information retrieval, there are some problems such as the lack of accurate expression of user query requests, the mismatch between document and query and query optimization. Focusing on these problems, we propose the query expansion method based on conceptual semantic space with deep learning, this hybrid query expansion technique include deep learning and pseudocorrelation feedback, use the deep learning and semantic network WordNet to construct query concept tree in the level of concept semantic space, the pseudo-correlation feedback documents are processed by observation window, compute the co-occurrence weight of the words by using the average mutual information and get the final extended words set. The results of experiment show that the expansion algorithm based on conceptual semantic space with deep learning has better performance than the traditional pseudo-correlation feedback algorithm on query expansion.     

基于概念图的用户兴趣查询扩展模型的研究

查询扩展是信息检索技术研究的一个重要组成部分。目前的查询扩展是基于统一的用户模型,没有考虑到用户的个人兴趣,这对查询扩展的精确度造成了一定的影响。分析了产生这种问题的原因,提出了基于概念图的用户兴趣扩展模型,通过该模型来有效提高查询扩展的精确度。实验显示,该方法能有效提高查询的查全率和查准率。     

基于语义的信息检索模型

由于查询与文档中词语的不匹配现象导致一些相关的文档不能被成功地检索出来,在信息检索的研究与实现中,这是影响检索效果的一个很关键的问题。把概念图和知网结合起来,提出对应的相关反馈算法,重新计算词项权重,利用向量空间模型和语义相似度进行语义检索,并给出了语义检索模型。实验结果显示该方法取得了良好的效果。     

RUBRIC: A System for Rule-Based Information Retrieval

A research prototype software system for conceptual information retrieval has been developed. The goal of the system, called RUBRIC, is to provide more automated and relevant access to unformatted textual databases. The approach is to use production rules from artificial intelligence to define a hierarchy of retrieval subtopics, with fuzzy context expressions and specific word phrases at the bottom. RUBRIC allows the definition of detailed queries starting at a conceptual level, partial matching of a query and a document, selection of only the highest ranked documents for presentation to the user, and detailed explanation of how and why a particular document was selected. Initial experiments indicate that a RUBRIC rule set better matches human retrieval judgment than a standard Boolean keyword expression, given equal amounts of effort in defining each. The techniques presented may be useful in stand-alone retrieval systems, front-ends to existing information retrieval systems, or real-time document filtering and routing.     

Development of technology for customer requirement-based reference design retrieval

Engineering design is a knowledge-intensive process, and includes conceptual design, detailed design, engineering analysis, assembly design, process design, and performance evaluation. Each task involves various aspects of knowledge and experience. Whether this knowledge and experience can be effectively shared is key to increasing product development capability and quality, and also to reducing the duration and cost of the development cycle. Therefore, offering engineering designers various query methods for retrieving engineering knowledge is one of the most important tasks in engineering knowledge management.The study develops a technology for customer requirement-based reference design retrieval to provide engineering designers with easy access to relevant design and associated knowledge. The tasks involved in this research include (i) designing a customer requirement-based reference design retrieval process, (ii) developing techniques related to the technology for customer requirement-based reference design retrieval, and (iii) implementing a customer requirement-based reference design retrieval mechanism. The retrieval process comprises the steps of customer requirement-based query, case searching and matching, and case ranking. The technology involves (1) a structured query model for customer requirement, (2) an index structure for historical design cases, (3) customer requirement-based case searching and matching mechanisms, (4) a customer requirement-based case ranking mechanism, and (5) a case-based representation of designed entities.     

A content-search information retrieval process based on conceptual graphs

An intelligent information retrieval system is presented in this paper. In our approach, which complies with the logical view of information retrieval, queries, document contents and other knowledge are represented by expressions in a knowledge representation language based on the conceptual graphs introduced by Sowa. In order to take the intrinsic vagueness of information retrieval into account, i.e. to search documents imprecisely and incompletely represented in order to answer a vague query, different kinds of probabilistic logic are often used. The search process described in this paper uses graph transformations instead of probabilistic notions. This paper is focused on the content-based retrieval process, and the cognitive facet of information retrieval is not directly addressed. However, our approach, involving the use of a knowledge representation language for representing data and a search process based on a combinatorial implementation of van Rijsbergen’s logical uncertainty principle, also allows the representation of retrieval situations. Hence, we believe that it could be implemented at the core of an operational information retrieval system. Two applications, one dealing with academic libraries and the other concerning audiovisual documents, are briefly presented.     

Scalable Retrieval and Mining With Optimal Peer-to-Peer Configuration

We proposed to utilize the scalable peer-to-peer network to perform the content-based image retrieval and mining, i.e, P2P-CBIRM. The decentralized unstructured P2P model with certain overheads, i.e., peer clustering and update procedures, is adopted to compromise with the structured one while still reserving flexible routing control when peers join/leave or network fails. The peer CBIRM engine is designed to utilize multi-instance query with multi-feature types to effectively reduce network traffic while maintaining high retrieval accuracy. It helps to enhance the knowledge discovery and image data mining capability. The proposed P2P-CBIRM system provides the scalable retrieval and mining function that the query scope and retrieval accuracy can be adaptively and progressively controlled. To improve the query efficiency (recall-rate/query-scope), it effectively utilizes both: 1) forwarding query message (forward phase) to reduce the query scope and 2) transmitting retrieval results (backward phase) such that activated peers keep filtering high similarity images on the link-path toward the query peer. Experiments show that the query efficiency of the scalable retrieval approach is better than previous methods, i.e., firework query model and breadth-first search. It provides a scalable knowledge discovery platform for efficient image data mining applications. We also proposed to optimally configure the P2P-CBIRM system such that, under a certain number of online users, it would yield the highest recall rate. Simulations demonstrate that, with the optimal configuration, recall rates can be improved to 2.5 to 3 times larger while the network traffic of each peer is reduced to 30% of the original, under the same number of on-line users.