Comparing PPGIS and LBSN Data to Measure Emotional Perception of the City

• Published in: Digital Transformation and Global Society pp. 223 - 234
• Main Authors: Nenko, Aleksandra, Petrova, Marina
• Format: Book Chapter
• Language: English
• Published: Cham Springer International Publishing
• Series: Communications in Computer and Information Science
• Subjects: Emotion analysis; Emotional geography; Google Places; Imprecity; LBSN; PPGIS
• ISBN: 9783030378578; 3030378578
• ISSN: 1865-0929; 1865-0937
• DOI: 10.1007/978-3-030-37858-5_18

Analysis of emotions has received recognition in urban studies as a mean to understand subjective quality of life. Availability of spontaneous user-generated online urban data generated by users in location based social networks broadens possibilities for such analysis as described in a number of studies. However the LBSN data is not shared deliberately by users and is not meant to be an expression of emotions, which makes its representativeness and validity questionable. Another source of data - public participation geo-information systems - helps to overcome these limitations however may have its own, such as a small and biased sample. In this paper the results of the comparative analysis of the distribution of emotions in St. Petersburg, Russia, visualized with LBSN and PPGIS data, are presented. The dataset is formed from user-generated comments on urban venues from Google Places and data from PPGIS platform Imprecity (www.imprecity.ru), where citizens deliberately share their emotions and comments about public spaces. The data samples contain 1800 emotional marks from Imprecity and 2450 geolocated comments from Google Places marked by experts and then processed with Naïve Bayes Classifier. Comparison of positive and negative emotional maps created for Imprecity and Google Places shows shared tendencies in emotional distribution, such as concentration of emotions in the city centre and collocation of positive and negative emotions. There are also differences in emotional distribution: PPGIS data shows local “emotional” islands, which correspond to pedestrian areas and green spaces. The comparative analysis appears to be insightful and capable of revealing recurring spatial tendencies in subjective perception of the city.