Using latent semantic analysis to enhance the comprehensibility of hypertext systems.

Abstract

Hypertext documents are pieces or nodes of information connected to each other through hyperlinks, which is the basis for more complex learning technologies as e-learning platforms or electronic encyclopaedias. Because of its extensive use, there is a need for procedures and techniques that help authors to enhance the comprehensibility of hypertext. This paper presents a number of cognitive models that explain how learners navigate and comprehends hypertext documents. The CoLiDeS and CoLiDeS+ models describe how people navigate when they try to find specific information, while the Hypertext Comprehension Model addresses knowledge acquisition when their goal is to comprehend hypertext contents. Some problems predicted by these models that affect the comprehensibility of hypertexts are discussed. In addition, a number of procedures are described that uses latent semantic analysis (LSA) free-text tools in order to assess and improve their usability and comprehensibility.