“…For instance, social media data that contain a large volume of texts and images are becoming a major descriptive indicator of people's cognition and emotional feelings attached to urban place in terms of sign-in, re-emergence, repetition, and aggregation (Jenkins, Croitoru, Crooks, & Stefanidis, 2016;Wu, Wang, Shi, Gao, & Liu, 2019). Machine learning and data mining methods have become mainstream tools for extracting thematic topics from textual data to quantify human activities and emotions at specific places (Gao et al, 2017;Hu, 2018;Tran, Shin, & Choi, 2018). Popular approaches include the Word2Vec model and its geo-enhanced variants (Yan, Janowicz, Mai, & Gao, 2017;Zhai et al, 2019).…”