Location-Based Social Networking for Spatio-Temporal Analysis of Human Mobility

Location-Based Social Networking for Spatio-Temporal Analysis of Human Mobility

P. Shanthi Saravanan (J. J. College of Engineering and Technology, Tiruchirappalli, India), Sabin Deori (Researcher, Sivasagar, India), and Balasundaram S. R. (National Institute of Technology, Tiruchirappalli, India)
Copyright: © 2021 |Pages: 20
DOI: 10.4018/978-1-7998-7756-1.ch003
OnDemand PDF Download:
No Current Special Offers


Human location tracking and analysis have always been important domains with a wide range of implementations in areas such as traffic prediction, security, disaster response, health monitoring, etc. With the availability and use of GPS-enabled devices, it has become easier to obtain location traces of an individual. There may be situations when the current location may be difficult to trace due to device failures or any unforeseen situation. This is one reason why geo tagged social networking or LBSN (location-based social networking) data research is gaining popularity. This kind of geo-tagged data when collected over time from a crowd can be analyzed for various mobility patterns of the population. This chapter focuses on how to predict the location of the people during mobility. A sample study to predict the geographical location and points of interest of a user is explained with the help of random forest classifier. Also, the chapter highlights the security and privacy concerns when LBSN is used for human mobility analysis.
Chapter Preview


Mobility is an essential aspect of humans for the sake of fulfilling their needs for survival. Depending on the individuals the mobility factor may be viewed for any kind of study. People go to different places for jobs, for food, for purchase, for visiting places etc. Mobility relates to geography and human needs. With the tremendous growth of information and communication technologies, it becomes possible to link the physical world with the digital space thereby providing ample opportunities to understand and provide services to the moving people (Roth et al. 2011; Lenormand et al., 2014; Song et al., 2006; Abou-zeid et al., 2013). It becomes possible to extract knowledge on human location details with the help of GPS, wearable devices, etc. This knowledge helps to monitor health aspects, seamless computing, disaster management, traffic prediction, etc. Much research can be seen in modeling and understanding human mobility patterns based on individual levels to groups pertaining to varieties of applications. Broadly speaking, understanding the patterns of mobility and applying them in transportation planning, resource allocation, spreading of diseases(Ravenstein1885; Siminiet al., 2012; Merler&Ajelli, 2010; Marguta&Parisi, 2015).

Many questions are possible with respect to human mobility. How far an individual or group of people will travel? Whether the mobility follows patterns? Whether time and space parameters contribute to the study of human mobility? Is it possible to know the future locations of humans based on past history? Other than location based details, by any other means, whether the location prediction of human mobility be tracked.

Hess et al. (2015) focused on classifying human mobility models into two groups namely trace based and synthetic(Hess et al., 2015).Trace based models are based on GPS, Cell Detail Records etc. Whereas synthetic models are outcomes of simulations based on mathematical computations. When technologies help to extract the location data of individuals to analyze the movements, data coming through social interactions also help to model the human mobility.

These models are predominantly observed in varieties of domains at general level such as people walking in a constrained space, walking to markets or any other business locations, etc. to more specific as well as broader domains such as urban planning, epidemic studies, traffic predictions etc. Lazer et al. (2009) specify digital traces as records of human activities collected through and stored by means of digital devices. Lima, Antonio (2016) et al. focus on models related to digital traces that include large number of forms of data gathered from devices and related societies.

In case of Calling Description Records (CDRs), generated from telephone exchange, the time, location, duration, and position of base stations, source and destination number etc. that are associated with each phone call could be used to predict the mobility(Horak, 2012).Whenever a call is done by a user, the nearest cellular network tower routing that call is recorded. This in turn yields the user’s geographical location.

Complete Chapter List

Search this Book: