Identifying activity-travel points from GPS-data with multiple moving windows

Dijk, Justin van

doi:10.1016/j.compenvurbsys.2018.02.004

Cited by 32 publications

(9 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They have been particularly useful in resource‐poor environments, although the vast number of mobile phone users provides an opportunity for tracking movement patterns of people throughout both the developed and developing world (Chen et al, ; Searle et al., ; Vazquez‐Prokopec et al., ; Wesolowski et al., ). Despite the popularity of mobile phone data in human mobility and transportation research (Alessandretti, Sapiezynski, Lehmann, & Baronchelli, ; Brum‐Bastos, Long, & Demšar, ; Chen et al., ; Feng & Timmermans, ; Gong, Chen, Bialostozky, & Lawson, ; Schneider, Belik, Couronné, Smoreda, & González, ; Sila‐Nowicka et al, ; Van Dijk, ; Williams, Thomas, Dunbar, Eagle, & Dobra, ), studies on the accuracy or the coverage of location data are relatively sparse. Here it is worth noting that the present paper focuses on “active mobile phone data” in which the location of the mobile phone is determined in response to queries specifically designed to collect location data at fixed time intervals or distance thresholds (Ahas, Aasa, Roose, Mark, & Silm, ; Sagl, Delmelle, & Delmelle, ).…”

Section: Introductionmentioning

confidence: 99%

Quality of hybrid location data drawn from GPS‐enabled mobile phones: Does it matter?

Yoo

Roberts

Eum

et al. 2020

Transactions in GIS

View full text Add to dashboard Cite

Despite their increasing popularity in human mobility studies, few studies have investigated the geo‐spatial quality of GPS‐enabled mobile phone data in which phone location is determined by special queries designed to collect location data with predetermined sampling intervals (hereafter “active mobile phone data”). We focus on two key issues in active mobile phone data—systematic gaps in tracking records and positioning uncertainty—and investigate their effects on human mobility pattern analyses. To address gaps in records, we develop an imputation strategy that utilizes local environment information, such as parcel boundaries, and recording time intervals. We evaluate the performance of the proposed imputation strategy by comparing raw versus imputed data with participants’ online survey responses. The results indicate that imputed data are superior to raw data in identifying individuals’ frequently visited places on a weekly basis. To assess the location accuracy of active mobile phone data, we investigate the spatial and temporal patterns of the positional uncertainty of each record and examine via Monte Carlo simulation how inaccurate location information might affect human mobility pattern indicators. Results suggest that the level of uncertainty varies as a function of time of day and the type of land use at which the position was determined, both of which are closely related to the location technology used to determine the location. Our study highlights the importance of understanding and addressing limitations of mobile phone derived positioning data prior to their use in human mobility studies.

show abstract

Section: Introductionmentioning

confidence: 99%

Quality of hybrid location data drawn from GPS‐enabled mobile phones: Does it matter?

Yoo

Roberts

Eum

et al. 2020

Transactions in GIS

View full text Add to dashboard Cite

show abstract

“…With respect to performance improvement, how features are extracted can significantly affect the results of travel‐mode classification. For the last decade, some studies considered the focal characteristics of sub‐segments for each trip, captured through moving windows sliding on GPS trajectories (Bolbol, Cheng, Tsapakis, & Haworth, ; Dodge, Weibel, & Forootan, ; van Dijk, ; Xiao et al, ). Bolbol et al () applied a fixed‐size moving window sliding on speed and acceleration values of multi‐segment GPS instances, and the classification of six travel modes using support vector machine (SVM) achieved an accuracy of 88%.…”

Section: Past Studies On Pa and Transport‐mode Classificationmentioning

confidence: 99%

“…Dodge et al () and Xiao et al () especially adopted the focal characteristic of GPS trajectories by calculating movement parameters from the GPS points that fall within a sliding window, achieving a predictive accuracy of 82% and 91%, respectively. van Dijk () achieved over 99% predictive accuracy in classifying trips (moving) and activities (staying) by introducing moving spatial and temporal windows.…”

Section: Past Studies On Pa and Transport‐mode Classificationmentioning

confidence: 99%

“…This study, in particular, uses GPS trajectory operators extensively at the focal level with different sizes of moving spatial and temporal windows, as described in Section . Compared to the research by van Dijk () that used spatial and temporal moving windows (as shown in Table ), this study extracts more features to maximize the benefit of spatial and temporal moving windows for classifying more travel modes by utilizing two different kinds of open sensor data.…”

Section: Past Studies On Pa and Transport‐mode Classificationmentioning

confidence: 99%

See 1 more Smart Citation

Automatic physical activity and in‐vehicle status classification based on GPS and accelerometer data: A hierarchical classification approach using machine learning techniques

Lee

Kwan

2018

Transactions in GIS

View full text Add to dashboard Cite

Due to the advancement of tracking technology, a large quantity of movement data has been collected and analyzed in various research domains. In human mobility and physical activity (PA) research, GPS trajectories and the capabilities of geographic information systems (GIS) facilitate a better understanding of the associations between PA and various environmental factors taking individuals' daily travels into account. PA research, however, needs to widen its focus from the intensity of PA to types of PA, which may provide useful clues for understanding specific health behaviors in particular geographic contexts. This study proposes and develops an algorithm to automatically classify PA types and in-vehicle status using GPS and accelerometer data. Walking, standing, jogging, biking and sedentary/in-vehicle statuses are identified through hierarchical classification processes based on machine learning and geospatial techniques. The proposed algorithm achieved high predictive accuracy on real-world GPS and accelerometer data. It can greatly reduce participants' and researchers' burdens by automatically identifying PA types and in-vehicle status for human mobility research, which is also known as travel mode imputation in transportation research. | 1523 LEE and KWAN

show abstract

“…trip ends and employ geographic information system (GIS) data to detect trip purposes [5]- [9]. By contrast, transportation modes are detected with either rule-based [10]- [12] or machine learning classification methods [13]- [16].…”

Section: Introductionmentioning

confidence: 99%

Detecting Travel Modes Using Rule-Based Classification System and Gaussian Process Classifier

2019

View full text Add to dashboard Cite

Travel modes are generally derived from Global Positioning System (GPS) data on the basis of either a rule-based or machine learning classification method. The rule-based classification approach is generally easy to understand, whereas the machine learning classification method has better generalization. However, studies that jointly explore both methods are limited. The present research proposes a two-stage method that aims to impute travel modes from GPS trajectory data. In the first stage, rules are employed to detect subway modes. In the second stage, a Gaussian process classifier based on sequential forward selection methods is utilized to derive the remaining travel modes. On the basis of the four selected features constituting the feature set (i.e., average speed, average acceleration, heading change, and low-speed point rate), over 97% of the samples with subway modes are correctly identified and 93.04% of segments in the walk-based balanced test subset are accurately detected. Over 92% of the car and bus samples are correctly detected for the training and test datasets. Results provide a new perspective in selecting classification methods for the detection of travel modes and other travel characteristics from GPS trajectory data. Furthermore, high differentiation is achieved between the bus and car modes without the bus transit geographic information system sources of bus networks. Therefore, reasonable extracted features contribute to the detection of travel modes, particularly between bus and car modes. INDEX TERMS Gaussian processes, classification algorithms, global positioning system.

show abstract

Identifying activity-travel points from GPS-data with multiple moving windows

Cited by 32 publications

References 22 publications

Quality of hybrid location data drawn from GPS‐enabled mobile phones: Does it matter?

Quality of hybrid location data drawn from GPS‐enabled mobile phones: Does it matter?

Automatic physical activity and in‐vehicle status classification based on GPS and accelerometer data: A hierarchical classification approach using machine learning techniques

Detecting Travel Modes Using Rule-Based Classification System and Gaussian Process Classifier

Contact Info

Product

Resources

About