感受性视角下人智交互研究:主题挖掘、方法识别、未来发展  

Human-AI Interaction Research from a Receptivity Perspective:Topic Classification,Method Identification,and Future Development

作　　者：陈静[1] 张璐[1] Chen Jing

机构地区：[1]华中师范大学信息管理学院,湖北武汉430079

出　　处：《情报理论与实践》2024年第2期98-106,共9页Information Studies:Theory & Application

基　　金：国家自然科学基金面上项目“APP交互环境下移动搜索行为机理研究”(项目编号:72074094);华中师范大学中央高校基本科研业务费项目“开源跨模态科技情报知识组织与智能分析”(项目编号:CCNU22QN016)的成果。

摘　　要：[目的/意义]人工智能研究正经历从人工智能主义到人本人工智能主义的视角转变。将两视角看作两“桥墩”,感受性因强调人和人工智能代理之间的关系过程,成为架起通向更准确、高效人工智能理解的“桥梁”。以感受性为视角,系统梳理以人智交互为对象的相关研究主题、方法与发展具有重要意义。[方法/过程]以现有国内外文献为基础,借助开放式编码从人智需求识别和人智交互体验两方面以及研究方式、研究方法论、研究方法三方面进行研究主题和方法回顾,继而对感受性视角下人智交互的未来研究方向进行展望。[结果/结论]一方面,以人智能力和信息需求识别为研究主题,利用定性研究方法探索人和人工智能代理双方的认知感受性;另一方面,以人智能力差异和人智体验评估为研究主题,借助定量研究方法量化人和以用户为中介的人工智能代理的感知感受性。[创新/局限]从研究工具升级、研究主题扩展、研究要素嵌套三维度来提升现有研究框架;以定性描述为主,未来需进一步加强定量分析。[Purpose/significance]AI research is undergoing a process whereby a perspective is shifting from artificial intelligence to human-centered AI.Treating them as two“bridge piers”can built bridge toward a more effective understanding,since receptivity emphasizes relationship process between human and agents.This study aims to systematize research on human-AI interaction from a receptivity perspective.[Method/process]Based on the existing literature,this study reviewed research topics from need identification and interaction experience of human and agents,and methods from research approach,research methodology and research methods by open coding.Then,the study prospected the future research direction.[Result/conclusion]On the one hand,qualitative method was used to explore cognition receptivity of human and agents from identification of ability and information needs.On the other hand,quantitative method was used to quantify perception receptivity of human and human-mediated from ability difference and interaction experience.[Innovation/limitation]Enhancing the existing research framework from three perspectives:research tool upgrading,research topic expanding and research element adding.This study focused on qualitative description,and further quantitative analysis should be conducted.

关 键 词：人智交互 感受性 人本人工智能主义 通用人工智能 人智需求识别 人智体验评估 

分 类 号：G353.1[文化科学—情报学]