PLoS ONE (Jan 2023)

Efficacy improvement in searching MEDLINE database using a novel PubMed visual analytic system: EEEvis

  • Jong-Chan Lee,
  • Brian J. Lee,
  • Changhee Park,
  • Hyunjoo Song,
  • Chan-Young Ock,
  • Hyojae Sung,
  • Sungjin Woo,
  • Yuna Youn,
  • Kwangrok Jung,
  • Jae Hyup Jung,
  • Jinwoo Ahn,
  • Bomi Kim,
  • Jaihwan Kim,
  • Jinwook Seo,
  • Jin-Hyeok Hwang

Journal volume & issue
Vol. 18, no. 2

Abstract

Read online

PubMed is the most extensively used database and search engine in the biomedical and healthcare fields. However, users could experience several difficulties in acquiring their target papers facing massive numbers of search results, especially in their unfamiliar fields. Therefore, we developed a novel user interface for PubMed and conducted three steps of study: step A, a preliminary user survey with 76 medical experts regarding the current usability for the biomedical literature search task at PubMed; step B is implementing EEEvis, a novel interactive visual analytic system for the search task; step C, a randomized user study comparing PubMed and EEEvis. First, we conducted a Google survey of 76 medical experts regarding the unmet needs of PubMed and the user requirements for a novel search interface. According to the data of preliminary Google survey, we implemented a novel interactive visual analytic system for biomedical literature search. This EEEvis provides enhanced literature data analysis functions including (1) an overview of the bibliographic features including publication date, citation count, and impact factors, (2) an overview of the co-authorship network, and (3) interactive sorting, filtering, and highlighting. In the randomized user study of 24 medical experts, the search speed of EEEvis was not inferior to PubMed in the time to reach the first article (median difference 3 sec, 95% CI -2.1 to 8.5, P = 0.535) nor in the search completion time (median difference 8 sec, 95% CI -4.7 to 19.1, P = 0.771). However, 22 participants (91.7%) responded that they are willing to use EEEvis as their first choice for a biomedical literature search task, and 21 participants (87.5%) answered the bibliographic sorting and filtering functionalities of EEEvis as a major advantage. EEEvis could be a supplementary interface for PubMed that can enhance the user experience in the search for biomedical literature.