Validating Sentiment Analysis on Opinion Mining Using Self-reported Attitude Scores

This paper aims to validate the use of sentiment analysis of text on opinion mining, using IBM Watson's Natural Language Understanding. An online survey examining attitude towards genomic editing using a validated scale and a text entry to comment on their thoughts of this technology was given to four populations-scientists, policymakers, farmers and the general public. The results indicate that sentiment scores of texts are significantly related to attitude scores in three of the four populations. This study shows that sentiment analysis is a reliable tool to understand opinions.