Ecemiş A.Dokuz A.Ş.Çelik M.2019-08-012019-08-0120199.78154E+12https://dx.doi.org/10.1109/IDAP.2018.8620832https://hdl.handle.net/11480/15212018 International Conference on Artificial Intelligence and Data Processing, IDAP 2018 -- 28 September 2018 through 30 September 2018 -- -- 144523Socially important locations are frequently visited locations of social media users in their social media lifetime. Socially important locations reveal spatial preferences of social media users in a social media user group. Discovering socially important locations is important for recommending locations, marketing, urban planning, etc. However, the motivation of preferring these locations is unclear. This study is performed to reveal the motivation of location preferences of social media users. Polarity of socially important locations of social media users are detected using two popular document classification methods namely, SVM and Naïve Bayes. The results showed that SVM algorithm is more efficient than Naïve Bayes algorithm on detecting tweet polarities. Also, the location preference motivations could be revealed based on dominant polarity ratio. © 2018 IEEE.eninfo:eu-repo/semantics/closedAccessNaïve Bayespolarity detectionSocial media miningsocially important locationsSVMTwitterConference Object10.1109/IDAP.2018.86208322-s2.0-85062537176N/AWOS:000458717400110N/A