JOURNAL OF CHINA UNIVERSITIES OF POSTS AND TELECOM ›› 2017, Vol. 24 ›› Issue (1): 67-76.doi: 10.1016/S1005-8885(17)60189-4

Previous Articles     Next Articles

Social media mining and visualization for point-of-interest recommendation

Ren Xingyi, Song Meina, E Haihong, Song Junde   

  1. 1. School of Computing, Beijing University of Posts and Telecommunications 
    2. Engineering Research Center of Information Networks, Beijing University of Posts and Telecommunications
  • Received:2016-09-05 Revised:2016-12-19 Online:2017-02-28 Published:2017-02-28
  • Contact: Xing-Yi REN
  • Supported by:
    This work was supported by the National Key Project of Scientific and Technical Supporting Programs of China (2014BAH26F02).

Abstract: With the rapid growth of location-based social networks (LBSNs), point-of-interest (POI) recommendation has become an important research problem. As one of the most representative social media platforms, Twitter provides various real-life information for POI recommendation in real time. Despite that POI recommendation has been actively studied, tweet images have not been well utilized for this research problem. State-of-the-art visual features like convolutional neural network (CNN) features have shown significant performance gains over the traditional bag-of-visual-words in unveiling the image’s semantics. Unfortunately, they have not been employed for POI recommendation from social websites. Hence, how to make the most of tweet images to improve the performance of POI recommendation and visualization remains open. In this paper, we thoroughly study the impact of tweet images on POI recommendation for different POI categories using various visual features. A novel topic model called social media Twitter-latent Dirichlet allocation (SM-TwitterLDA) which jointly models five Twitter features, (i.e., text, image, location, timestamp and hashtag) is designed to discover POIs from the sheer amount of tweets. Moreover, each POI is visualized by representative images selected on three predefined criteria. Extensive experiments have been conducted on a real-life tweet dataset to verify the effectiveness of our method.

Key words: social media, Twitter, POI recommendation, visualization

CLC Number: