Tuning the matching function for a threshold weighting semantics in a linguistic information retrieval system

Abstract

Information retrieval is an activity that attempts to produce documents that better fulfill user information needs. To achieve this activity an information retrieval system uses matching functions that specify the degree of relevance of a document with respect to a user query. Assuming linguistic-weighted queries we present a new linguistic matching function for a threshold weighting semantics that is defined using a 2-tuple fuzzy linguistic approach~Herrera F, Martínez L. We show that it simplifies the processes of computing in the retrieval activity, avoids the loss of precision in final results, and, consequently, can help to improve the users' satisfaction.

A computer-supported learning system to help teachers to teach Fuzzy Information Retrieval Systems

“…Each user query is expressed as a combination of the weighted terms which are connected by the logical operators AND ð^Þ; OR ð_Þ; and NOT ð:Þ: The weights associated with the query terms could be numerical values assessed in [0, 1] or linguistic values taken from a linguistic term set S defined in a fuzzy ordinal linguistic context (Herrera and Herrera-Viedma 2000;Herrera-Viedma 2001a, b;Herrera-Viedma et al 2005) (see Appendix 1). In this context, a user query is any legitimate Boolean expression whose atomic components (atoms) are pairs \t i ; w i [ ; t i 2 T and being w i 2 I; I 2 ½0; 1 or I 2 S the weight associated to the term t i by the user.…”

“…It offers students the opportunity to see and compare the achieved results of different weighted queries. Student can choose (i) different semantics (threshold, relative importance, ideal importance, quantitative) (Herrera-Viedma 2001b;Herrera-Viedma et al 2005) to formulate weighted queries, (ii) different fuzzy connectives to evaluate such queries (maximum, minimum, Ordered Weighted Averaging (OWA) operators, Induced OWA operators, Linguistic OWA operators, and Linguistic Weighted Averaging operators) (Chiclana et al 2004;Chiclana et al 2007;Herrera and Herrera-Viedma 1997;Herrera et al 1996;Yager 1987Yager , 1988Yager and Filev 1999), and (iii) different expression domains (numerical and fuzzy ordinal linguistic one) (Herrera and HerreraViedma 2000;Herrera-Viedma 2001b; to assess weights associated with queries. Furthermore, several standard test collections (ADI, CISI, CRANSFIELD, TREC, etc.)…”

A Review on Information Accessing Systems Based on Fuzzy Linguistic Modelling

“…The fuzzy linguistic IRSs based on a 2-tuple FLM that have been defined are the following: 90 , 91 , and 92 .…”

“…In 91 Enrique Herrera-Viedma and Antonio G. López-Herrera and Carlos Porcel propose a fuzzy linguistic IRS based on the 2-tuple FLM that supports weighted queries based on a new interpretation of the symmetrical threshold semantics defined in 85 . The use of the 2-tuple FLM allows defining a new matching functions that improves the interpretation of the symmetrical threshold semantics proposed in 85 .…”

“…Many fuzzy linguistic IRSs 17,85,86,91,92 uses symmetrically and uniformly distributed linguistic term sets in their retrieval activity (see Figure 3). In these cases, the same discrimination levels on both sides of the middle linguistic term are established.…”