Mobility prediction using movement, semantic and social data: a survey

Abstract

I. INTRODUCTIONData mining is one of the steps for future prediction and knowledge discovery. Due to the rapid development of technology with respect to software, hardware and automation systems, huge amount of data has been generated and stored in data marts and data warehouses. The traditional informal statistical methods and other such tools used to mine data are not sufficient to analyse this huge collection of data. Hence, an intelligent data analysis technique known as Knowledge discovery has been identified, so that knowledge or patterns can be extracted from various data sets. The mobility prediction is defined as tracking the movement patterns of the human or object and predicting their future locations based on the mined patterns. It can be categorized into two types: Trip-based, where aggregated mobility is considered (trips between different areas are predicted from the recorded data, and source destination matrices are generated); Activity-based, where individual mobility is considered (each individual entity is locations to access, and based on their actions, trips are generated) [1]. Mobility prediction has attracted good research interest due to its vast applications like prediction of single or sequence of locations, mobility pattern behaviour for group or individual, urban service based applications, etc., It could be applied to any field such as government sectors, healthcare units, scientific domains and other similar domains, so that they are able to discover the patterns and predict results well in advance to safeguard the collapses which might occur in future. The best wa y to collect the movement data is through the GPS enabled devices especially mobile phones [2]. In Wireless Sensor Networks (WSN), the object mobility prediction has a variety of benefits like smart homes, enterprise applications, environmental monitoring, education and military hospital services [3]. In this work, we present a survey of movement prediction of human and objects based on next single or sequence of locations. The remainder of the paper is organized as follows. Section II discusses the background, motivation and basic definitions. Section III presents the study on various applications on mobility predictions of both human and object. The comparative assessment of the study is presented in Section IV and Section V concludes the paper with future works. II. BACKGROUND AND MOTIVATIONPredicting the future move of human or object helps us to reserve resources. Some of such movement prediction applications are Vehicle routing problem, social networking applications, reminder, aged people movement optimization, recommender systems and assistive devices. These applications are beneficial to the common people, commercial organizations, government agencies, health care units and safety critical systems. The definitions of the movement data such as trajectory, temporal, spatial, semantic and social data are given below:  A trajectory is the path made by the moving entity through the space where it moves [4].  Temporal...