作为搜索引擎的关键技术之一，查询推荐研究正受到越来越多的关注。由于查询推荐技术与应用领域联系紧密，方法不具有一般普适性，因此研究需要针对不同领域数据的不同特征采用相对应的合适方法对其进行处理。本文以专业知识与工艺流程等为研究对象，以盐湖化工领域为例，在领域本体的基础上利用有向带权图重新组织其概念关系，并将其划分为不同知识包，构建了一个面向工艺流程的专业领域知识库。在领域知识库的基础上本文提出一种新的查询推荐算法，实验结果表明该算法有效地提高面向专业领域的主题检索质量与精度。%As one of the essential technologies for search engine, query recommendation is attracting more and more attention. Query recommendation technology is relevant with the application field so closely that the method is not universal, so the related researches should adopt appropriate method to deal with different kinds of information in accordance with their characteristics. This article set expertise and process flow as the research objects, utilized the weighted directed graph to re organize the concepts which are based on the domain ontology, and divided the graph into different knowledge packages, which constituted the knowledgebase. Based on the knowledgebase, a new query recommendation algorithm was put forward, whose quality and accuracy for retrieval in professional field was demonstrated by the experimental results.
展开▼
