近年来,实验室大力开展心理健康计算领域的研究,将社交媒体作为观察人们心理健康状态的窗口,利用心理学、认知科学、语言学、自然语言处理等理论和技术,对于人的性格、幽默、幸福感、阴郁症、自杀倾向等进行识别和预测。
近日,杨亮老师关于计算性格学方面的研究成果被Springer旗下计算机期刊Soft Computing录用,其为CCF推荐C类期刊。
题目:Computational Personality: a survey(计算性格学综述)
摘要:Personality is a set of stable and tendentious behaviors, thoughts and emotions. How to measure personality more conveniently and accurately has always been a problem for scholars in related fields. With the rapid development of computer technology and the widespread popularity of social media in recent years, the research of computational personality has attracted wide attention of researchers in Computational Linguistics and psychology. Various methods, from statistical methods in psychology to machine learning and then to deep learning, have been proposed to deal with different areas of computational personality. In this paper, we first summarize the research framework of computational personality, and then review the current research progress of computational personality from the aspects of personality prediction, depression detection, suicide detection and happiness assessment, and provide the corresponding research resources for reference. Finally, we provide some possible research directions.
性格是一组稳定的、倾向性的行为、思想和情感。如何更方便、准确地测量性格一直是相关领域学者聚焦的难题。近年来,随着计算机技术的飞速发展和社交媒体的普及,计算性格学的研究引起了计算语言学和心理学研究者的广泛关注,研究方案发展经历了统计方法到机器学习,再到深度学习。本文首先概述了计算性格学的研究框架,然后从性格预测、抑郁检测、自杀检测和幸福感评估等方面综述了计算性格学的研究现状,并提供了相应的研究资源供参考,最后提出了一些未来可能的研究方向。