Quantifying the Intensity of Toxicity for Discussions and Speakers
Quantifying the Intensity of Toxicity for Discussions and Speakers
复制标题
DOI：
10.1109/aciiw52867.2021.9666258
复制
发表时间：
2021-09
期刊：
影响因子：
--
通讯作者：
Samiha Samrose;E. Hoque
中科院分区：
文献类型：
--
作者：
Samiha Samrose;E. Hoque
在这项工作中，我们从 YouTube 新闻节目多模态数据集中与二元发言者进行了激烈的讨论，通过视听信号分析毒性。首先，由于不同的说话者对毒性的贡献可能不同，我们提出了按说话者进行的毒性评分，以揭示个体的成比例贡献。由于存在分歧的讨论可能反映出一些毒性信号，为了识别需要更多关注的讨论，我们将讨论分为高低毒性级别。通过分析视觉特征，我们发现这些水平与面部表情相关，因为上眼睑抬起（与“惊讶”相关）、酒窝（与“轻蔑”相关）和唇角压低（与“厌恶”相关）在区分高低不敬强度方面仍然具有统计显着性。其次，我们研究了基于音频的特征（例如音调和强度）可能显着引起不尊重的影响，并利用这些信号通过应用逻辑回归模型对不尊重和非不尊重样本进行分类，准确率达到 79.86%。我们的研究结果揭示了利用视听信号为理解有毒讨论添加重要背景的潜力。
In this work, from YouTube News-show multimodal dataset with dyadic speakers having heated discussions, we analyze the toxicity through audio-visual signals. Firstly, as different speakers may contribute differently towards the toxicity, we propose a speaker-wise toxicity score revealing individual proportionate contribution. As discussions with disagreements may reflect some signals of toxicity, in order to identify discussions needing more attention we categorize discussions into binary high-low toxicity levels. By analyzing visual features, we show that the levels correlate with facial expressions as Upper Lid Raiser (associated with ‘surprise’), Dimpler (associated with ‘contempť), and Lip Corner Depressor (associated with ‘disgust’) remain statistically significant in separating high-low intensities of disrespect. Secondly, we investigate the impact of audio-based features such as pitch and intensity that can significantly elicit disrespect, and utilize the signals in classifying disrespect and non-disrespect samples by applying logistic regression model achieving 79.86% accuracy. Our findings shed light on the potential of utilizing audio-visual signals in adding important context towards understanding toxic discussions.