Title :
Idea density — A potentially informative characteristic of retrieved documents
Author :
Covington, Michael A.
Author_Institution :
Inst. for Artificial Intell., Univ. of Georgia, Athens, GA, USA
Abstract :
Idea density, or number of propositions divided by number of words, is a well-known psycholinguistic measurement which can now be estimated reliably by software. Preliminary tests indicate that idea density distinguishes between documents about the same subject written for specialist and nonspecialist audiences, and that it does not correlate with lexical diversity or Flesch-Kincaid readability.
Keywords :
information retrieval; Flesch-Kincaid readability; document retrieval; idea density; potentially informative characteristics; psycholinguistic measurement; Artificial intelligence; Density measurement; Humans; Information retrieval; Logic; Production; Psychology; Software measurement; Speech; Testing;
Conference_Titel :
Southeastcon, 2009. SOUTHEASTCON '09. IEEE
Conference_Location :
Atlanta, GA
Print_ISBN :
978-1-4244-3976-8
Electronic_ISBN :
978-1-4244-3978-2
DOI :
10.1109/SECON.2009.5174076