The Mandarin Chinese Auditory Emotions Stimulus Database: A Validated Set of Subject Personal Pronoun Sentences (MCAE-SPPS)

When expressing emotions in daily life, sentences with different personal pronouns as the subject may bring listeners different emotional experiences. This study established and validated an auditory emotional speech dataset consisting of sentences with three types of Mandarin Chinese Personal Pronouns (I and We, singular You and plural You, singular third-person, and They) as the subject, an unprecedented resource. Six professional Mandarin actors (three men and three women) expressed two hundred meaningful Chinese sentences in neutral and six basic emotional tones: happiness, sadness, anger, fear, disgust, and surprise. Seven hundred and twenty Chinese college students evaluated the 8400 recordings for emotional categories (a one out of seven-force choose task) and intensity (a scale of 1-9). The final dataset consists of 7579 valid Chinese subject pronoun sentences (Neutrality: 1608; Sadness: 1324; Anger: 1167; Surprise: 1076; Disgust: 797; Happiness: 855; Fear: 752; First/Second/Thirds personal pronouns sentences: 3085, 2975, and 1519). Recognition rates for the emotions were as follows: neutrality (82%), sadness (80%), anger (78%), surprise (72%), happiness (65%), disgust (62%), and fear (60%). We provided each sound’s raw data, validation results, perceptual intensity rating scores, and acoustic information. This personal pronoun emotional sentence database is valuable for various research areas, including linguistics, psychology, neuroscience, clinical rehabilitation, and computer science. The Chinese subject personal pronouns auditory emotional speech database is at open science framework

在日常生活中表达情绪时，以不同人称代词为主语的语句可能会给听者带来不同的情绪体验。本研究构建并验证了一款听觉情绪语音数据集，该数据集以三类现代汉语人称代词（Mandarin Chinese Personal Pronouns）为主语语句为核心内容，涵盖我/我们、单数你/复数你们、单数第三人称与他/她/它们三类，是一项前所未有的研究资源。六名专业普通话演员（三名男性、三名女性）录制了200句富有意义的汉语语句，涵盖中性情绪及六种基本情绪语调：快乐、悲伤、愤怒、恐惧、厌恶与惊讶。720名中国大学生对8400条语音录音开展了情绪类别评估（七选一强制选择任务）与强度评分（1-9分量表）。最终数据集包含7579条有效的汉语人称代词语句：中性1608条、悲伤1324条、愤怒1167条、惊讶1076条、厌恶797条、快乐855条、恐惧752条；其中第一人称代词语句3085条、第二人称代词语句2975条、第三人称代词语句1519条。各情绪的识别准确率如下：中性（82%）、悲伤（80%）、愤怒（78%）、惊讶（72%）、快乐（65%）、厌恶（62%）与恐惧（60%）。本研究提供了每条语音的原始数据、验证结果、感知强度评分与声学信息。该人称代词情绪语句数据库在语言学、心理学、神经科学、临床康复与计算机科学等多个研究领域均具有重要应用价值。本汉语人称代词听觉情绪语音数据集已发布于开放科学框架（Open Science Framework）。

搜集汇总

背景与挑战

背景概述

该数据集是一个包含以不同人称代词为主语的中文句子，由专业演员以中性及六种基本情感语调表达的听觉情感语音数据库。数据集经过大量学生评估，提供了详细的音频数据、情感识别率和声学信息，适用于多学科研究。