# Beyond writing machines: A Kano model analysis of researchers’ hierarchical needs for AIGC services across the research lifecycle

## TL;DR

This study identifies and categorizes researchers' needs for AI-generated content (AIGC) tools across the research lifecycle to guide better service design and adoption.

## Contribution

A novel hierarchical framework of AIGC service needs using the Kano model and Importance-Performance Analysis for academic service optimization.

## Key findings

- Three must-be attributes include data security, citation accuracy, and reliability.
- Seven one-dimensional attributes include automated literature summarization and language polishing.
- Five attractive attributes include generating novel hypotheses and smart journal recommendations.

## Abstract

The proliferation of AI-Generated Content (AIGC) tools presents both opportunities and challenges for the academic service ecosystem. However, a systematic understanding of researchers’ multifaceted demands for AIGC functionalities remains underdeveloped, hindering the strategic design and optimization of these services. This study addresses this gap by investigating three core questions: (1) What specific AIGC service functions do researchers desire across the research lifecycle? (2) How can these needs be categorized hierarchically? (3) What is their relative importance in influencing user satisfaction? Employing an exploratory sequential mixed-methods design, this research first identified a comprehensive list of 15 service demands through semi-structured interviews with 45 expert researchers (N = 45). Subsequently, these demands were prioritized through a large-scale questionnaire survey involving 412 researchers (N = 412), utilizing the Kano model and Importance-Performance Analysis. The results reveal a clear hierarchy of needs: we identified three must-be attributes (e.g., data security, citation accuracy), seven one-dimensional attributes (e.g., automated literature summarization, language polishing), and five attractive attributes (e.g., generating novel research hypotheses, smart journal recommendation). These findings provide a detailed framework for AIGC service development and offer an evidence-based model for academic institutions to prioritize resource allocation, thereby enhancing the value and adoption of AIGC in scholarly research.