Automatic credibility assessment of popular medical articles available online

Abstract

This paper presents the design concept of a credibility evaluation tool for medical web-documents and describes the implementation of its part. There have been numerous attempts to create such tool but most of them were strictly subject-specific. In this study, we aim to create a universal classifier for non-credible articles from the medical domain. Unlike most of the latest fact-checking solutions, it evaluates overall the credibility of the document instead of assessing separate claims. We collected a database of articles and sentences evaluated by experts, conducted the study of sentence’s context in the task of credibility assessment, then performed statistical analysis in order to verify and fine-tune the design. The proposed scheme is constructed in such a way that it should be easy to update and has an easily interpretable output for Internet users with no expert knowledge about medicine.