Scholarlink

<p>This study proposes a complete and comprehensive solution for Product Innovation Topic Discovery (PITD) through online reviews, which accurately identifies user needs and expectations by topic mining and Aspect-Based Sentiment Analysis (ABSA). In this solution, a simplified algorithm, LGCF-BERT, is proposed for the ABSA component based on BERT’s local and global context-focused mechanism. Its effectiveness was verified on Chinese online review datasets. An improved quantitative KANO model is used to classify all topics in a more objective manner and determine the product innovation topics. Further, this solution demonstrates its feasibility and practicality in a real business environment.</p>

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

A refined approach of product innovation topic discovery through online reviews based on aspect-based sentiment analysis and KANO model

  • 期刊论文

  • 发表日期：

    2025年8月20日

查看全文
  • Yuanfang Dong,
  • Jia Song,
  • Ruowei Li

摘要

This study proposes a complete and comprehensive solution for Product Innovation Topic Discovery (PITD) through online reviews, which accurately identifies user needs and expectations by topic mining and Aspect-Based Sentiment Analysis (ABSA). In this solution, a simplified algorithm, LGCF-BERT, is proposed for the ABSA component based on BERT’s local and global context-focused mechanism. Its effectiveness was verified on Chinese online review datasets. An improved quantitative KANO model is used to classify all topics in a more objective manner and determine the product innovation topics. Further, this solution demonstrates its feasibility and practicality in a real business environment.