Sentiment analysis and audience engagement in thehashtags #VcNoJL1 and #VcNoJL2 on Twitter (X): a casestudy of the TV Liberal newscast in Pará, Brazil

Abstract

Television broadcasts around the world employ specific hashtags to facilitate viewer engagement via social media with program hosts. In the context of news broadcasts, individuals' comments mix with the articles shown on television news, mainly in relation to urban situations. These social media posts constitute a significant database, as users assume roles similar to those of citizens. The aim of this paper is to scrutinize the posts associated with the hashtags #VcNoJL1 and #VcNoJL2, publishe on the Twitter (X) platform, in the context of the TV news program TV Liberal, broadcast in the State of Pará (Brazil), using the CRISP-DM methodology and data science techniques. Using a statistical examination of these publications, a discreet level of engagement was observed among viewers, who are also users of the aforementioned social media platform. Furthermore, a sentiment assessment of the posts was carried out, revealing that 54% of the comments are negative, 31% are positive and 15% are neutral. The negative comments notably highlight the challenges within municipal public administration, while the positive comments are around commemorative occasions such as Mother's Day and Grandparents' Day. This type of analysis constitutes an important monitoring tool not only for the television program, but also for public administration.

Keywords

• Television news
• Hashtags
• Live interaction
• Data science
• Sentiment analysis