FuFaIR: a Fuzzy Farsi Information Retrieval System

Abstract

Persian (Farsi) is one of the languages of Middle East. There are significant amount of Persian documents available in digital form and even more are created every day. Therefore, there is a necessity to implement Information Retrieval System with high precision for this language. This paper discusses the design, implementation and testing of a Fuzzy retrieval system for Persian called FuFaIR. This system also supports Fuzzy quantifiers in its query language. Tests have been conducted using a standard Persian test corpus called Hamshari. The performance results obtained from FuFaIR are positive and they indicate that the FuFaIR could notably outperform well known industry systems such as the vector space model.

A Distributed N-Gram Indexing System to Optimizing Persian Information Retrieval

“…Fuzzy logic [19], VSM [20] and language modeling methods [21]were used in the studies done about the retrieval of Persian documents. Persian test collection is introduced in [7] where authors used a number of documents in law domain.…”

“…Also experiments in [21] suggest the usefulness of language modeling techniques for Persian. Oroumchian et al [20] used N-gram method in search for Persian documents and they give better results by applying LCA (Local Context Analyses) and N-Grams. Considering trigrams and 4-grams extracted from documents and not using stemming process they showed that compared with VSM, 4-gram method increases retrieval efficiency for searching.…”

“…Persian language is one of the used languages in the Middle East, so there are significant amount of Persian documents available on the web [16,20]. There are 32 letters in Persian language.…”

Experiments with English-Persian text retrieval

“…The authors in [10] proposed the design and testing of a Fuzzy retrieval system for Persian (FuFaIR) with support of Fuzzy quantifiers in its query language and their experiments showed that the retrieval system outperforms the vector space model. Also experiments in [2] Furthermore, Compression of Persian text for web was assessed on Hamshahri collection in [12] in which the authors could obtain up to 52% reduction in size.…”

“…Where w m,n is calculated using equation (10). Then Combinatorial Translation Probability (CTP) is a |T(q 1 )|×|T(q M )| matrix that is calculated by multiplication of all of the W k matrices:…”