Effects of terminology on health queries: An analysis by user's health literacy and topic familiarity

• Type: Article in International Scientific Journal
• Date: 2015

Authors

Carla Lopes (Author) FEUP; Cristina Ribeiro (Author) FEUP

Indexing

Scopus - 12 Citations

Scientific classification

CORDIS: Physical sciences > Computer science > Informatics

FOS: Natural sciences > Computer and information sciences

Other information

Authenticus ID: P-00G-C11

DOI: 10.1108/s0065-283020150000039013

Abstract (EN): Prior studies have shown that terminology support can improve health information retrieval but have not taken into account the characteristics of the user performing the search. In this chapter, the impact of translating queries' terms between lay and medico-scientific terminology, in users with different levels of health literacy and topic familiarity, is evaluated. Findings demonstrate that medico-scientific queries demand more from the users and are mostly aimed at health professionals. In addition, these queries retrieve documents that are less readable and less well understood by users. Despite this, medico-scientific queries are associated with higher precision in the top-10 retrieved documents results and tend slightly to generate knowledge with less incorrect contents, the researchers concluded that search engines should provide query suggestions with medico-scientific terminology, whenever the user is able to digest it, that is, in users above the lowest levels of health literacy and topic familiarity. On the other hand, retrieval systems should provide lay alternative queries in users with inadequate health literacy or in those unfamiliar with a topic. In fact, the quantity of incorrect contents in the knowledge that emerges from a medico-scientific session tends to decrease with topic familiarity and health literacy. In terms of topic familiarity, the opposite happens with Graded Average Precision. Moreover, users most familiar with a topic tend to have higher motivational relevance with medico-scientific queries than with lay queries. This work is the first to consider user context features while studying the impact of a query processing technique in several aspects of the retrieval process, including the medical accuracy of the acquired knowledge. © 2015 by Emerald Group Publishing Limited.

Language: English

Type (Professor's evaluation): Scientific