Comparing discrimination powers of text and citation-based context types

Hajar Sotudeh () and Mojgan Houshyar ()
Additional contact information
Hajar Sotudeh: Shiraz University
Mojgan Houshyar: Shiraz University

Scientometrics, 2018, vol. 114, issue 1, No 10, 229-251

Abstract: Abstract Different context types have different potentialities in representing subjects of documents. However, little is known on their probable differences regarding discriminating documents with different subjects. The present study aimed to compare the discrimination powers of five context types (i.e. title, textual, text-citation, reference, and reference-title contexts). Using a content analysis method, a test collection consisted of 3637 papers containing 15 English homographs in 54 subject clusters were analyzed by discriminant analysis to determine the discrimination among the clusters. The results confirmed that the homograph context types significantly discriminate between the clusters. The contexts can correctly cluster the documents from about 40 to 80%, with text and reference-title context types being stronger in differentiating the clusters. This is the first study to compare the powers of different textual contexts kinds and recommend to concentrate on context types to discriminate between documents with homographs to improve the effectiveness of their retrieval. It is suggested that a comparison of the context types be used as a tool in re-ranking or clustering of the retrieved results.

Keywords: Discriminant analysis; Title context; Reference-title context; Text-citation context; Textual context; Reference context (search for similar items in EconPapers)
Date: 2018
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
http://link.springer.com/10.1007/s11192-017-2566-9 Abstract (text/html)
Access to the full text of the articles in this series is restricted.

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:spr:scient:v:114:y:2018:i:1:d:10.1007_s11192-017-2566-9

Ordering information: This journal article can be ordered from
http://www.springer.com/economics/journal/11192

DOI: 10.1007/s11192-017-2566-9

Access Statistics for this article

Scientometrics is currently edited by Wolfgang Glänzel

More articles in Scientometrics from Springer, Akadémiai Kiadó
Bibliographic data for series maintained by Sonal Shukla () and Springer Nature Abstracting and Indexing ().