Case Knowledge Retrieval Algorithm And Applied Research

In daily life, people often need to retrieve the similar historical experience and knowledge to be the reference of solving current problems. And the experience and methods in human cognitive process mostly saved as text. So, textual case retrieve has become the demand of information times.This paper mainly studies textual case retrieve algorithms and applications, the main works are as follows:1) The representation of textual case. According to the particle granularity characteristics of the textual case, analysis the granulation of textual case refer to the process of information granulation, use the strategy of observing the problem from different levels, combined with the characteristics of human understanding mode which is using the sentence as a unit, the sentence vector space model is carried out to quantify the expression of textual case. The expression size of textual case is increased from word to sentence, and simple semantic information is considered.2) Textual case retrieve algorithms. The process of textual case retrieve is abstracted as granular computing through the analysis of granularity principle in textual case retrieve, and the textual case retrieve algorithms based on sentence vector space model is proposed, we confirmed that the algorithm is feasible and improving the search results through experiment.3) Improvement of textual case retrieve algorithms. First, redundancy in textual case impacts the retrieval efficiency and speed, so the key sentences idea is proposed using the 80/20 rule and used to improve the textual case retrieve algorithms, and the experiment shows that the algorithm increases the retrieval speed while improves the search results. Second, domain knowledge is considered in process of textual case retrieve, the idea of constructing the domain keywords library is proposed, the textual case retrieve algorithms is improved again. The experiment proves that the method is feasible.4) Increased speed of retrieval. Textual case retrieval algorithm parallelization is proposed because the case knowledge base is so large that the retrieval takes too long time. The parallel computing is realized based on MPI parallel computing platform, and increased speed of retrieval. 5) The implementation of textual case retrieval system. Based on the core academic ideas of this paper and the design idea, we introduce the overall structure and workflow, and describe the design and implementation of main modules. At last, prototype system is given to verify the model and algorithm proposed in this paper.