Similarity based Retrieval of Documents using Jaccard Coefficient Similarity Methods

Particularly, information retrieval resultsas documents are typically too extensive.Consequently, a similarity measurement betweenkeywords and index term is essentially performedto facilitate searchers in accessing the requiredresults. Thus, this paper proposed the similaritymeasurement method between words by deployingJaccard Coefficient. After documents arecollected, some pre-processing tasks such as stopwordremoval, stemming, and handling of digits,hyphens, punctuation, and cases of letters areusually performed. Almost 200 documents (withpdf formats) from UCSYare conducted in thissystem. With the help of this system, the users cansearch effectively the most related documents withtheir preference information without boring.Furthermore, the performance of the proposedsimilarity measurement method: Jaccardsimilarity coefficient was accomplished byemploying precision, recall, and F- measure