MLModeler5 at SemEval-2023 Task 3: Detecting the Category and the Framing Techniques in Online News in a Multi-lingual Setup

Proceedings of the The 17th International Workshop on Semantic Evaluation (SemEval-2023), pages 1096–1101 July 13-14, 2023 ©2023 Association for Computational Linguistics MLModeler5 at SemEval-2023 Task 3: Detecting the Category and the Framing Techniques in Online News in a Multi-lingual Setup Arjun Khanchandani, Nitansh Jain, Jatin Bedi Computer Science and Engineering Department Thapar Institute of Engineering and Technology arjunkhanchandani@gmail.com, njain_be20@thapar.edu, jatin.bedi@thapar.edu Abstract Throughout the ages, we have seen various leaders, dictators, politicians, CEOs, compa- nies, sports agencies, organizations like the UN harness the power of news articles to shape and influence the minds of the people of the world. Various persuasion and framing tech- niques are employed while writing articles to achieve this aim. It has become necessary now, more so than ever, to identify these techniques and strategies to help understand these articles, reduce the spread of fake news, stop propa- ganda stories and ensure that people of the world are well informed. To help achieve these tasks, we have implemented RoBERTa and AL- BERT models for Subtask 1/2 Task 3 SemEval- 2023. Both these models are BERT-based ar- chitectures. The results of these models are validated on the dataset provided by the orga- nizers of this task. 1 Introduction Popular mass media comes together and shapes the perceptions of people and communities on numerous subjects of our society [1]. Moni- toring mass media in different languages and understanding and analyzing niche as well as broad domain topics is essential in this day and age. With the increase in the amount of data be- ing generated and processed, and the proliferation of news as well as information around us in recent times, it is essential to perform in-depth comparisons of how data is being put in front of the public by different mass media of different countries. There is also a need to promote and support the automation of this media analysis process [10]. The main drive behind this task is to foster the development of methods and tools to support the analysis of online media content in order to understand what makes a text persuasive: which writing style is used, what critical elements are highlighted, and which persuasion techniques are used to influence the reader’s mind and thinking. There is an essential need to identify and develop methods that identify what techniques are being used in media content. “The term persuasion comes from the Latin word persuasion, which could be translated as an act of persuading, convincing. Persuasion is a process, where the main motive is talking a recipient round; convincing, mentally influencing, guiding, or directing someone from one place to another.” [5]. In a polarized media environment, partisan media outlets intentionally frame news stories in a way to advance specific political agendas[7]. “Even when journalists make their best efforts to pursue objectivity, media framing often favors one side over another in political disputes, thus always resulting in some degree of bias” [4]. Hence, a news framing analysis is helpful because it not only tells us whether a news …

Content truncated to 500 words