The enormous use of sarcastic text in all forms of communication in social media will have a physiological effect on target users. Each user has a different approach to misusing and recognising sarcasm. Sarcasm detection is difficult even for users, and this will depend on many things such as perspective, context, special symbols. So, that will be a challenging task for machines to differentiate sarcastic sentences from non-sarcastic sentences. There are no exact rules based on which model will accurately detect sarcasm from many text corpus in the current situation. So, one needs to focus on optimistic and forthcoming approaches in the sarcasm detection domain. This paper discusses various sarcasm detection techniques and concludes with some approaches, related datasets with optimal features, and the researcher's challenges.
翻译:社交媒体中讽刺文本的广泛使用将对目标用户产生生理影响。每个用户在误用和识别讽刺方面都有不同的方法。讽刺检测对用户来说甚至也很困难,这将取决于许多因素,例如视角,上下文,特殊符号等。因此,机器区分讽刺性句子和非讽刺性句子将是一项具有挑战性的任务。在当前情况下,没有精确的规则可以使模型从大量文本语料库中准确检测出讽刺。因此,人们需要关注讽刺检测领域中乐观且前瞻性的方法。本文讨论了各种讽刺检测技术,并以一些方法,相关数据集以及研究者面临的挑战作出了结论。