Scholarlink

In this paper we focus upon building on the existing methods of using persona and auxiliary information on improving and humanizing a conversational model as there are still limitations when it comes to personification and dialogue generation by incorporating the users’ linguistic proficiency with the CEFR (Common European Framework of Reference for Languages) categories. OneStopEnglish corpus dataset and Readability dataset are used in the neural network models to predict the CEFR levels. The user-provided persona information is used in the Large Language Models (LLMs) to provide auxiliary information to personify and generate dialogues and scripts. In addition to advancing dialog models, this multidisciplinary study establishes a professional benchmark for enhancing user engagement in the ever-changing world of over-the-top, movies, series, novels and anime-inspired fan content.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Persona-Driven Dialog Generation: Enhancing User Engagement Through Linguistic Proficiency and Personalization

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • A. Abhishek,
  • R. Akhilesh,
  • B. Swetha,
  • V. Vindhya Vasini,
  • T. Deepika

摘要

In this paper we focus upon building on the existing methods of using persona and auxiliary information on improving and humanizing a conversational model as there are still limitations when it comes to personification and dialogue generation by incorporating the users’ linguistic proficiency with the CEFR (Common European Framework of Reference for Languages) categories. OneStopEnglish corpus dataset and Readability dataset are used in the neural network models to predict the CEFR levels. The user-provided persona information is used in the Large Language Models (LLMs) to provide auxiliary information to personify and generate dialogues and scripts. In addition to advancing dialog models, this multidisciplinary study establishes a professional benchmark for enhancing user engagement in the ever-changing world of over-the-top, movies, series, novels and anime-inspired fan content.