基于映射的信息检索逻辑模型研究

Research on the Imaging-based Information Retrieval Logic Model

本文讨论了基于逻辑映射的信息检索模型,对该模型进行实例分析,指出该模型在实际应用中存在的一些问题。这些问题包括各可能世界的先验概率分布和各可能世界间的相似关系的计算。文中分析了当不同的文档中都包含有查询中的标引词时,如何区分文档的重要性,并对其中一些问题提出了改进的方法。

The paper discusses the information retrieval model based on logic imaging, illustrates the model by examples and points out its problems in actual applications. The problems involve the computation of the prior probability distribution for each possible world and the similarity between possible worlds. The paper describes how to distinguish the importance of documents in the case that different documents contain the terms in the query. It also proposes some measures for improvement.