Scholarlink

This paper presents a novel approach to enhancing user experience (UX) in automated environments through Aspect-Based Sentiment Analysis (ABSA), a subfield of Natural Language Processing (NLP). With advanced AI technologies, ABSA analyzes text for general sentiment and extracts and evaluates sentiments tied to specific aspects or features discussed in the text. Our research aims to evaluate the effectiveness of ABSA as a tool for informed UX decision-making in design processes. In this collaborative effort, we compiled a dataset of 8,060 tweets from 2016 to 2023 using keywords related to automation and smart workplaces. After rigorous pre-processing, we utilized structured tweets for aspect extraction. Six different pre-trained or finetuned ABSA models were employed to analyze these tweets, focusing on their utility in refining UX. Our findings highlight how ABSA can be leveraged to enhance user interactions. We propose a set of strategic recommendations to help researchers and decision-makers identify strengths and weaknesses in their automated systems. Ultimately, we present an organizational life cycle model that utilizes sentiment classification to refine UX strategies, facilitating better AI integration in user-centric designs.

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

Enhancing User Experience in Automated Systems Using Aspect-Based Sentiment Analysis

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Parisa Saadati,
  • Laden Husamaldin,
  • Favour Aladesuru,
  • Jose Abdelnour Nocera

摘要

This paper presents a novel approach to enhancing user experience (UX) in automated environments through Aspect-Based Sentiment Analysis (ABSA), a subfield of Natural Language Processing (NLP). With advanced AI technologies, ABSA analyzes text for general sentiment and extracts and evaluates sentiments tied to specific aspects or features discussed in the text. Our research aims to evaluate the effectiveness of ABSA as a tool for informed UX decision-making in design processes. In this collaborative effort, we compiled a dataset of 8,060 tweets from 2016 to 2023 using keywords related to automation and smart workplaces. After rigorous pre-processing, we utilized structured tweets for aspect extraction. Six different pre-trained or finetuned ABSA models were employed to analyze these tweets, focusing on their utility in refining UX. Our findings highlight how ABSA can be leveraged to enhance user interactions. We propose a set of strategic recommendations to help researchers and decision-makers identify strengths and weaknesses in their automated systems. Ultimately, we present an organizational life cycle model that utilizes sentiment classification to refine UX strategies, facilitating better AI integration in user-centric designs.