Next place prediction using mobility Markov chains

Gambs, Sébastien; Killijian, Marc-Olivier; Núñez-del-Prado, Miguel

doi:10.1145/2181196.2181199

Cited by 403 publications

(277 citation statements)

References 12 publications

Supporting

Mentioning

267

Contrasting

Unclassified

Order By: Relevance

“…A number of studies [7,8] have established that the second-order Markov model (2-MM) has the best accuracies, up to 95%, for predicting human mobility, and that higher-order MM (>2) is not necessarily more accurate, but is often less precise. However, the 2-MM always utilizes historical geo-spatial trajectories to train a transition probability matrix and in 2-MM (see Figure 3a) the probability of each destination is computed based only on the present and immediate past grids of interest that a user visited without using temporal information.…”

Section: Second-order Markov Model For Trajectory Predictionmentioning

confidence: 99%

“…TraPlan contains three essential techniques: (1) constrained network R-tree (CNR-tree), which is a two-tiered dynamic index structure of moving objects based on transportation networks; (2) a region-of-interest (RoI) discovery algorithm, which is employed to partition a large number of trajectory points into distinct clusters; and (3) a Trajectory-Prediction (TP) approach based on frequent trajectory patterns (FTP) tree, called FTP-mining, which is proposed to discover FTPs to infer future locations of objects within RoIs. The Markov chain (MC) model has been adopted by a number of works on predicting human mobility [7,8] to incorporate some amount of memory. Second-order MC has the best accuracies, up to 95%, for predicting human mobility, and higher order MC (>2) is not necessarily more accurate, but is often less precise.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Sparse Trajectory Prediction Based on Multiple Entropy Measures

Zhang

Liu

Xia

et al. 2016

Entropy

View full text Add to dashboard Cite

Trajectory prediction is an important problem that has a large number of applications. A common approach to trajectory prediction is based on historical trajectories. However, existing techniques suffer from the "data sparsity problem". The available historical trajectories are far from enough to cover all possible query trajectories. We propose the sparsity trajectory prediction algorithm based on multiple entropy measures (STP-ME) to address the data sparsity problem. Firstly, the moving region is iteratively divided into a two-dimensional plane grid graph, and each trajectory is represented as a grid sequence with temporal information. Secondly, trajectory entropy is used to evaluate trajectory's regularity, the L-Z entropy estimator is implemented to calculate trajectory entropy, and a new trajectory space is generated through trajectory synthesis. We define location entropy and time entropy to measure the popularity of locations and timeslots respectively. Finally, a second-order Markov model that contains a temporal dimension is adopted to perform sparse trajectory prediction. The experiments show that when trip completed percentage increases towards 90%, the coverage of the baseline algorithm decreases to almost 25%, while the STP-ME algorithm successfully copes with it as expected with only an unnoticeable drop in coverage, and can constantly answer almost 100% of query trajectories. It is found that the STP-ME algorithm improves the prediction accuracy generally by as much as 8%, 3%, and 4%, compared to the baseline algorithm, the second-order Markov model (2-MM), and sub-trajectory synthesis (SubSyn) algorithm, respectively. At the same time, the prediction time of STP-ME algorithm is negligible (10 µs), greatly outperforming the baseline algorithm (100 ms).

show abstract

Section: Second-order Markov Model For Trajectory Predictionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Sparse Trajectory Prediction Based on Multiple Entropy Measures

Zhang

Liu

Xia

et al. 2016

Entropy

View full text Add to dashboard Cite

show abstract

“…So the instability of the Wi-Fi signal leads to the indoor positioning as in [9] is more difficult than the outdoor positioning as in [8]. The second also the most important reason is [8] build an individual model to predict where the specified one person to go, but [9] still use one single model to do the general prediction among different persons. The original design of n-MMC is used to model individual mobile trace but not for the general predictive purpose of crowds.…”

Section: Related Workmentioning

confidence: 99%

“…Gambs et al [7], [8] proposed a Mobility Markov-chain Model (MMC) to incorporate the n previous visited locations and its extension coined as n-MMC to predict next possible [9] conducted another evaluation of n-MMC in the indoor context and the prediction rate is only up to 49% when n=2. Two main realistic context differences lead to n-MMC cannot reach the similar prediction accuracy in indoor context as the one in outdoor context.…”

Section: Related Workmentioning

confidence: 99%

“…Firstly, for Wi-Fi based IPS, it's easier that the Wi-Fi signal is reflected and absorbed when it encounters another medium with different electrical properties due to the limited room in which the signal propagates. So the instability of the Wi-Fi signal leads to the indoor positioning as in [9] is more difficult than the outdoor positioning as in [8]. The second also the most important reason is [8] build an individual model to predict where the specified one person to go, but [9] still use one single model to do the general prediction among different persons.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Wi-Fi Based Indoor Next Location Prediction Using Mixed State-Weighted Markov-Chain Model

Huang¹,

Dahlmeier²,

Lin³

et al. 2014

IJMLC

View full text Add to dashboard Cite

I. INTRODUCTIONNowadays, Wi-Fi enabled connectivity has become one of the standard functions for smart phones and tablets with the trend of increasing number of indoor Wi-Fi enabled venues. For example, shopping malls provide free Wi-Fi connection to attract more shoppers. These free Wi-Fi services can bring more traffic to increase the visit volume of the shopping malls. Wi-Fi is also a "must-have" for retail stores [1]. Retailers want to be the special preference of their customers, in order to do that, they desire to have a direct relationship with their customers and remember their preferences. For example, once the customers come to their shops, these customers can be identified via Wi-Fi signal which is captured by their mobiles or tablets, a popup will be displayed in their mobile devices to inform them about new products or promotions and provide recommendations based on their historical purchase record.In addition, in order to attract more customers into the shopping mall, these Wi-Fi services can also be used to help capturing indoor shoppers' location. Given a particular location, the smart phone can be represented by its Wireless LANs Address (MAC) uniquely and the received strength of the Wi-Fi signal indicator (RSSI) from different access points Manuscript received April 9, 2014; revised July 8, 2014 can be measured. By modeling the distribution of the RSSI in different locations, an Indoor Positioning System (IPS) as in [2], [3] can be built and trained with collected RSSI data to predict the device location from the signal strength. Although there are other options for indoor positioning, such as the indoor-messaging-system (IMES) in which positioning uses the indoor-GPS devices [4], however, because of the low-cost setup and the simple infrastructure needed, the Wi-Fi based IPS is more attractive to many retailers. Normally, the accuracy of these Wi-Fi Based IPS ranges from 3 to 10 meters [5]. If we keep recording the location of a MAC address for a period of time, these chronological location data can represent the movement or trajectory of a specified person.In this work, we aim to develop a model to predict the pedestrian's next location based on his historical movement which can be represented by his trajectory. There are many potential ways in which users can benefit from this application of the next location prediction. By being aware of the movement of the shoppers in advance, the retailers can quickly target shoppers by both interest and location, turning window-shopping into actual sales [6]. In order to maximize the effect of coupon promotion, delivering the coupons in specified regions and at a particular time, namely "user oriented coupon dispersion", can be carried out. Leveraging the application of next location prediction, organizers can rearrange the manpower with more efficiency and at a lower cost.This paper presents an improved model for the next location prediction and the remainder of this paper is organized as follows. First, we review relevant related works in Section II b...

show abstract

Correlated data in differential privacy: Definition and analysis

Zhang

Zhu

Liu

et al. 2020

Concurrency and Computation

View full text Add to dashboard Cite

Summary Differential privacy is a rigorous mathematical framework for evaluating and protecting data privacy. In most existing studies, there is a vulnerable assumption that records in a dataset are independent when differential privacy is applied. However, in real‐world datasets, records are likely to be correlated, which may lead to unexpected data leakage. In this survey, we investigate the issue of privacy loss due to data correlation under differential privacy models. Roughly, we classify existing literature into three lines: (1) using parameters to describe data correlation in differential privacy, (2) using models to describe data correlation in differential privacy, and (3) describing data correlation based on the framework of Pufferfish. First, a detailed example is given to illustrate the issue of privacy leakage on correlated data in real scenes. Then our main work is to analyze and compare these methods, and evaluate situations that these diverse studies are applied. Finally, we propose some future challenges on correlated differential privacy.

show abstract

Next place prediction using mobility Markov chains

Cited by 403 publications

References 12 publications

Sparse Trajectory Prediction Based on Multiple Entropy Measures

Sparse Trajectory Prediction Based on Multiple Entropy Measures

Wi-Fi Based Indoor Next Location Prediction Using Mixed State-Weighted Markov-Chain Model

Correlated data in differential privacy: Definition and analysis

Contact Info

Product

Resources

About