Developing recommendation techniques for location based social networks using random walk

---

Tezin Türü: Doktora

Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Bilgisayar Mühendisliği Bölümü, Türkiye

Tezin Onay Tarihi: 2015

Öğrenci: HAKAN BAĞCI

Danışman: PINAR KARAGÖZ

Özet:

The location-based social networks (LBSN) enable users to check-in their current location and share it with other users. The accumulated check-in data can be employed for the benefit of users by providing personalized recommendations. In this thesis, we propose three recommendation algorithms for location-based social networks. These are random walk based context-aware location (CLoRW), activity (RWCAR) and friend (RWCFR) recommendation algorithms. All the algorithms consider the current context (i.e. current social relations, personal preferences and current location) of the user to provide personalized recommendations. We propose an undirected unweighted graph model for representing LBSN data that contains users, locations and activities. We build a graph according to the current context of the user for each algorithm depending on this LBSN model. A random walk with restart approach is employed on this graph to predict the recommendation scores. Lists of users, locations and activities are recommended to users after ordering the nodes according to estimated scores. We compare CLoRW with popularity-based, friend-based and expert-based baselines, collaborative filtering approach and a similar work in the literature. According to results, our location recommendation algorithm outperforms these approaches in all of the test cases. Moreover, we also compare RWCAR and RWCFR algorithms with respective popularity-based, friend-based and expert-based baselines. In all of the experiments, RWCAR and RWCFR perform better than the baselines. The results clearly indicate that random walk based context-aware recommendation approach is a good candidate for recommending locations, activities and friends for LBSNs.