Quantifying the bias in place emotion extracted from photos on social networking sites: A case study on a university campus

《Quantifying the bias in place emotion extracted from photos on social networking sites: A case study on a university campus》

打印

作者单位: School of Resource and Environmental Sciences, Wuhan University, 129 Luoyu Road, Wuhan 430079, China;MNR Key Laboratory for Geo-Environmental Monitoring of Great Bay Area & Guangdong Key Laboratory of Urban Informatics & Shenzhen Key Laboratory of Spatial Smart Sensing and Services, Shenzhen University, No. 3688 Nanhai Avenue, Shenzhen 518060, China;School of Resource and Environmental Sciences, Wuhan University, 129 Luoyu Road, Wuhan 430079, China;MNR Key Laboratory for Geo-Environmental Monitoring of Great Bay Area & Guangdong Key Laboratory of Urban Informatics & Shenzhen Key Laboratory of Spatial Smart Sensing and Services, Shenzhen University, No. 3688 Nanhai Avenue, Shenzhen 518060, China

摘要: Various fields have widely used place emotion extracted from social networking sites (SNS) information in recent years. However, the emotional information may contain biases as users are a particular subset of the whole population. This research studies whether there are significant differences between place emotion extracted from SNS and the place in-situ (a campus of Wuhan University). Two datasets from different sources, Weibo (a platform similar to twitter) and in-situ cameras, are collected over the same time periods in the same geographical range. By utilizing online cognitive services on the photos collected, the diversity of people with a recognizable face in terms of age, gender, and emotions are determined. The results suggest that there are significant differences in place emotion extracted from Weibo and in-situ. Furthermore, the pattern of differences varies among diverse demographic groups. This paper quantitatively contrasts place emotion extracted from SNS and the place in-situ, which can help researchers achieve a more profound understanding of human behavior differences between online and offline place emotion. This research also provides a theoretical basis to calibrate the emotion metrics obtained from SNS facial expressions on future place emotion studies.