Real-Time Sentiment Analysis of Domestic Violence Tweets: Informing Intervention Strategies

Intimate partner violence is still a relevant topic in modern societies all over the world. which requires constant activities for the purpose of treatment and prevention. It is equally important to grasp attitudes of the public towards domestic violence as well as the overall attitude that is spread in the society. vital when it comes to identifying specific approaches to address the concerns and influence the population. discourse. Based on the context defined above, this paper aims to present a sentiment analysis solution. designed especially for the purpose of analyzing the posts that fall under the category of domestic. violence. By using the never-ending stream of data in the use of business applications, it is possible to increase accountability of employee productivity. For this purpose, we follow our method on Twitter to offer the corresponding insights into public. concerning this sensitive issue, the ideas and feelings people have towards it are as follows: topic. First of all, we gather a large amount of tweets containing tweets containing the relevant keywords about domestic violence employing the Twitter Application Programming Interface. Next, in order to manipulate the text data we have, we require processing it to remove noise, tokenizetion, and apply techniques such as stemming, lemmatization or both. to standardize the text. Subsequently, we employ stateof-the-art analysis - which is performed with the help of some specific algorithms for sentiment analysis maybe based on deep learning. models like, for instance, recurrent neural networks or transformers. architectures as BERT (Bidirectional Encoder Representations derived from Transformers to transcribe the tweet into a certain category of “positive”, “negative” or “neutral” sentiment. positive, negative, or neutral