Similarity model based on CBR and FCA

Chongyang Shi, Linjing Lai*, Jing Fan, Yu Bai

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Citations (Scopus)

Abstract

Case-based reasoning (CBR) is one of the research highlights in the artificial intelligence field. In the process of case retrieval of CBR, Similarity is an important index in evaluation. This paper proposed a new model calculating similarity between source case and target case. The model is suitable for using Formal Concept Analysis (FCA) in the case of CBR. The model considered the case attributes weights between two formal concepts and feature attributes weights in concept lattices. Comparing to the similarity model put forward by Jirapond Tadrat, this model cut down the comparison to the objects, weight more in the attributes and shorten the time of solving the similarity. Theoretical deduction proves that the proposed similarity model satisfy the basic conditions which all these models need to meet. This article chose the UCI data sets and the method of cross validation, carried on an experiment from both similarity model aspect and classifier aspect respectively. The former experimental results show that the similarity model has higher accuracy than others. The latter experimental results show that the similarity model of CBR classifier has higher accuracy in the attribute set density compared with other small data set classifier.

Original languageEnglish
Title of host publication2016 IEEE/ACIS 17th International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2016
EditorsYihai Chen
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages597-603
Number of pages7
ISBN (Electronic)9781509022397
DOIs
Publication statusPublished - 18 Jul 2016
Event17th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2016 - Shanghai, China
Duration: 30 May 20161 Jun 2016

Publication series

Name2016 IEEE/ACIS 17th International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2016

Conference

Conference17th IEEE/ACIS International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2016
Country/TerritoryChina
CityShanghai
Period30/05/161/06/16

Keywords

  • Case library
  • Case-based reasoning
  • Concept lattice
  • Formal concept analysis

Fingerprint

Dive into the research topics of 'Similarity model based on CBR and FCA'. Together they form a unique fingerprint.

Cite this

Shi, C., Lai, L., Fan, J., & Bai, Y. (2016). Similarity model based on CBR and FCA. In Y. Chen (Ed.), 2016 IEEE/ACIS 17th International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2016 (pp. 597-603). Article 7515965 (2016 IEEE/ACIS 17th International Conference on Software Engineering, Artificial Intelligence, Networking and Parallel/Distributed Computing, SNPD 2016). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/SNPD.2016.7515965