Public Transportation Alighting Estimation Method Using Smart Card Data

Kim, Dong Kyu; Lee, Inmook

doi:10.7782/jksr.2017.20.5.692

Cited by 6 publications

(11 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Based on this assumption, they estimate alighting stops when there are two or more boarding records per day for the same card ID [1][2][3][4][5]. Some studies assume that the passenger returns to the resident in the case of the last trip on the analyzed day, while the origin of the first trip on a pertinent day is the resident [7][8][9][10][11]. In some other cases, studies look at the origin of the first trip on the next day to estimate the alighting stop of the final trip on that day [12][13].…”

Section: Methodology For Estimating Destination Based On Trip Chainmentioning

confidence: 99%

“…Some studies validated the estimation results to improve the success rate and accuracy of alighting stop estimations and analyzed major parameters and sensitivity. Specifically, studies were conducted on the validation of trip-chain-based alighting stop estimation results [16][17]; sensitivity analysis for "allowed walking distance" to determine the continuity between two trips as a major parameter when estimating the alighting stop based on the trip-chain [9][10]; determining whether the passenger transferred and analyzing the walking journey time in terms of the trip-chain-based estimation [11,18].…”

Section: Methodology For Estimating Destination Based On Trip Chainmentioning

confidence: 99%

“…To determine the spatial continuity, we classified a trip as continuous travel, only when there was a stop on the used route of the classification-target trip within a certain distance from the boarding stop of the next trip. According to the result of sensitivity analysis on allowed walking distance in a previous study, the change in the success rate and accuracy of the estimation is not large when it is around 500m (Kim and Lee, 2017). In some cases, the allowed walking distance is set to 500 m, considering that the common radius of a subway station area is 500m, or it is set to 500-1,000 m based on the distance distribution between the bus stops in the analyzed area (Alsger et al, 2016, Nunes et al, 2016, Gordon et al, 2013, Yang et al, 2019.…”

Section: Classification Of Trip Typesmentioning

confidence: 98%

“…To overcome these limitations, various studies have been conducted in South Korea and other countries to estimate the alighting stops when the alighting tag information of users is missing. In particular, some studies have been conducted focusing on alighting stop estimation methodologies based on the connectivity between trips, i.e., under the assumption that the next trip starts around the destination of this trip [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18]. Furthermore, studies have been conducted on alighting stop estimation methods by applying machine learning methods and analysis of travel patterns based on the past travel history data [12][13][14][15][16].…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Estimating Destination of Bus Trips Considering Trip Type Characteristics

Lee¹,

Lee²,

Bae³

et al. 2021

Preprint

View full text Add to dashboard Cite

In recent years, local governments have been using transportation card data to monitor the use of public transport and improve the service. However, local governments that are applying a single-fare scheme are experiencing difficulties in using data for accurate identification of real travel patterns, policy decision support, etc. because the information on alighting stops of users is missing. This policy limits its functionality of utilizing data such as accurate identification of real travel patterns, policy decision support, etc. Various studies to overcome this limitation have been conducted in South Korea and other countries to develop es-timation methodologies of alighting stops. Even existing studies introduce an advanced method, we found the margin for better accuracy by combining various estimation methodologies for estimating alighting stops. This study reviewed previously conducted studies to classify data with missing alighting stop information into trip types and then applied an appropriate alighting stop estimation methodology for the characteristics of each trip type by stage. The proposed method is evaluated by utilizing transportation card data of the Seoul metropolitan area and checked the accuracy for each standard of allowable error for sensitivity analysis. Furthermore, the number of trips, accuracy, and valid tag rate were checked for each type to examine the need for classifying the trip types. Finally, our evaluation also examines the impact of classifying trip types on estimation accuracy. The evaluation criteria are accuracy of the number of trips and valid tag rate. The analysis shows that the stage-by-stage estimation methodology based on the trip type proposed in this study can es-timate users’ destinations more accurately than previous studies. Furthermore, based on the construction of nearly 100% valid tag data, this study differs from prior studies.

show abstract

Section: Methodology For Estimating Destination Based On Trip Chainmentioning

confidence: 99%

Section: Methodology For Estimating Destination Based On Trip Chainmentioning

confidence: 99%

Section: Classification Of Trip Typesmentioning

confidence: 98%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Estimating Destination of Bus Trips Considering Trip Type Characteristics

Lee¹,

Lee²,

Bae³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Although there may be differences in the behavior of omission of alighting stops due to differences in public transportation fare systems by country, the type of data collected and the approach for estimating alighting stops are similar. In particular, there are several studies focusing on alighting stop estimation methodologies based on the connectivity between trips, i.e., under the assumption that the next trip starts near the destination of this trip [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18]. Furthermore, studies on alighting stop estimation by applying machine learning methods and analyzing travel patterns based on the past travel history data have been conducted [12][13][14][15][16].…”

Section: Introductionmentioning

confidence: 99%

Estimating Destination of Bus Trips Considering Trip Type Characteristics

Lee

Bae

et al. 2021

Applied Sciences

View full text Add to dashboard Cite

Recently, local governments have been using transportation card data to monitor the use of public transport and improve the service. However, local governments that are applying a single-fare scheme are experiencing difficulties in using data for accurate identification of real travel patterns or policy decision support due to missing information on alighting stops of users. This policy limits its functionality of utilizing data such as accurate identification of real travel patterns, policy decision support, etc. In order to overcome these limitations, various methods for estimating alighting stops have been developed. This study classifies trips with missing alighting stop information into trip four types and then applies appropriate alighting stop estimation methodology for each trip type in stages. The proposed method is evaluated by utilizing transportation card data of the Seoul metropolitan area and checking the accuracy for each standard of allowable error for sensitivity analysis. The analysis shows that the stage-by-stage estimation methodology based on the trip type proposed in this study can estimate users’ destinations more accurately than the methodologies of previous studies. Furthermore, based on the construction of nearly 100% valid tag data, this study differs from prior studies.

show abstract