Amazing technological breakthrough possible @S-Logix

Office Address

  • #5, First Floor, 4th Street Dr. Subbarayan Nagar Kodambakkam, Chennai-600 024 Landmark : Samiyar Madam
  • +91- 81240 01111

Social List

Research Proposal in Deep Neural Network-based Sarcasm Detection with Multi-Task Learning

Research Proposal in Deep Neural Network-based Sarcasm Detection with Multi-Task Learning

  Sarcasm detection is the demanding sub-task of sentiment analysis and opinion mining applications in the field of natural language processing (NLP). Sarcasm detection identifies whether the text is sarcastic or non-sarcastic, expressed as negative attitudes with positive contrary words by the user. Determining the sarcasm is inadequate to achieve the interconnection with sentiment analysis.

  Exploiting multi-task learning in sarcasm detection assists in attaining robust interconnection with sentiment analysis. Multi-task learning focus on solving multiple tasks simultaneously based on commonalities and differences across tasks. Multi-task learning for sarcasm detection considers sentiment classification opinion analysis as stages of tasks to perform the main task as sarcasm detection. Multi-task learning utilizes deep learning models to extract powerful representations to be shared across different tasks. The effectiveness of the sarcasm detection model is enhanced by applying a deep neural network with multi-task learning.

  • Identifying sarcasm evinces a focused research field in Natural Language Processing (NLP), which is beneficial to avoiding misinterpretation of sarcastic statements as true statements.

  • Multi-task learning has gained popularity in the Artificial Intelligence (AI) communities, as it influences useful information in multiple related tasks to enhance the generalization performance of all the tasks.

  • Significant dominance of choosing multi-task learning are improved data efficiency, reduced overfitting through shared representations, and fast learning.

  • In sarcasm detection, multi-task detection helps to analyze the sarcastic sentences by assigning multiple tasks, including sentiment analysis and sarcasm detection, based on several factors of sentiment shifts.

  • Learning such multiple tasks imposes the deep neural networks to learn good representations to recognize sarcasm and sentimental shifts.

  • Multi-task learning with a deep neural network appreciably outperforms self-standing sarcasm analysis by permitting task interaction and knowledge sharing for both sentiment and sarcasm detection.