Healthcare Misinformation Detection and Fact-Checking: A Novel Approach

Information gets spread rapidly in the world of internet. The internet has become first choice people for medication tips related to their health problems. However, this ever-growing usage also led misinformation. misinformation healthcare severe effects on life people, thus efforts are required detect as well fact-check information before using it. In paper, authors proposed a model and domain. extracts healthcare-related URLs from web, pre-processes it, computes Term-Frequency, sentimental grammatical features misinformation, distance measures viz. Euclidean, Jaccard, Cosine similarity True or False based manually generated dataset with expert’s opinions. was evaluated five state-of-the-art machine learning classifiers Logistic Regression, Support Vector Machine, Naïve Bayes, Decision Tree, Random forest. experimental results showed that crucial while detecting more negative words found containing compared having true information. It observed Bayes outperformed all other models terms accuracy showing 98.7% whereas decision tree classifier less an 92.88%. Also, Jaccard Distance measure be best algorithm Euclidean measures.