Skip to content
BY-NC-ND 3.0 license Open Access Published by De Gruyter March 7, 2017

An Extensive Review on Data Mining Methods and Clustering Models for Intelligent Transportation System

Sesham Anand, P. Padmanabham, A. Govardhan and Rajesh H. Kulkarni


Data mining techniques support numerous applications of intelligent transportation systems (ITSs). This paper critically reviews various data mining techniques for achieving trip planning in ITSs. The literature review starts with the discussion on the contributions of descriptive and predictive mining techniques in ITSs, and later continues on the contributions of the clustering techniques. Being the largely used approach, the use of cluster analysis in ITSs is assessed. However, big data analysis is risky with clustering methods. Thus, evolutionary computational algorithms are used for data mining. Though unsupervised clustering models are widely used, drawbacks such as selection of optimal number of clustering points, defining termination criterion, and lack of objective function also occur. Eventually, various drawbacks of evolutionary computational algorithm are also addressed in this paper.

1 Introduction

In intelligent transportation systems (ITSs), developing a travel demand model is crucial [11, 13, 54, 55]. It not only facilitates transportation but also reduces greenhouse gas emissions [19]. Few examples include soft transport policy measures [6, 17] and the prediction of spatial and temporal emissions [19]. The traditional trip-based approach to study the travel demand fails due to its insensitivity toward the changing transportation policies and expanding infrastructures and services. Despite the fact that the activity-based approaches can handle insensitivity issues [41], the persisting uncertainty among the transportation systems poses great challenge [64].

Data mining approaches, especially clustering algorithms, can handle such uncertainties effectively [46, 61, 62]. As the effectiveness of data mining relies on data acquisition and exploration [35], the contributions are diverse. They range from constructing knowledge from the acquired transportation information [36], mining information [29], predicting [10], and planning transportation services as well as infrastructures [14, 53]. Hence, clustering methods support both the modeling and prediction of the characteristics of transportation [8].

This paper presents an extensive review on various data mining as well as clustering methods that model, predict, and plan transportation systems to facilitate ITSs. Around 50 research articles from the last decade were collected from the leading journals and reviewed in three stages (Figure 1). Firstly, the contributions that pertain to the ITS were reviewed chronologically. Secondly, the data mining approaches, involving predictive and descriptive concepts for managing transportation, were reviewed based on the adopted methodologies. In the third stage, the clustering models were categorized as supervised or unsupervised, and reviewed based on their usefulness over transportation data analysis. Eventually, the paper discusses our findings from the review and summarizes the research gaps and future directions of research.

Figure 1: Statistical Report of Research Articles.

Figure 1:

Statistical Report of Research Articles.

2 Review of Transportation Management System

2.1 Chronological Review

In 2007, Jarasuniene [27] focused on ITS-enabling technologies to study the involvement of transport professionals in human factor experts at the starting phase of ITS equipment design. Huang et al. [24] proposed a new search approach with its heuristics based on the changes of the A* algorithm-Lifelong Planning A*. An ellipse has pruned the unnecessary nodes from being scanned. Liang et al. [40] developed a real-time approach to detect the cognitive distraction by using driving performance data and the driver’s eye movements. Xia and Chen [60] used the data clustering methodology to study the characteristics of traffic flow and to analyze the basic relationships among the traffic parameters.

In 2009, Byon et al. [9] developed a novel method to gather information of the travelers’ transportation modes in both the peak and the non-peak periods by tracking the Global Positioning System (GPS)-equipped mobile devices in the traffic hub. Crainic et al. [12] improved the final performance of freight ITS using a better operations research-based decision-support software system. Andreas et al. [2] studied the use of mobile-aided positioning approaches in GSM networks by a traffic congestion estimation service. The means of accuracy improvement in the known cellular positioning techniques has been found to launch a new traffic congestion estimation service application. Lee et al. [35] created a knowledge-dependent travel time prediction system that works in real time with data mining schemes to support the urban network. The two predictors’ dynamic weight combination using meta-rules has enhanced the precision of predicting the travel time.

In 2010, Jacob and Abdulhai [26] mitigated non-recurring and recurring congestion with an adaptive integrated freeway-arterial corridor control that is capable of self-learning. Lee et al. [36] proposed a joint structure for accumulating, fusing, as well as sharing traffic data in real time. It involved the user-centric traffic event reacting methodology and automatic agent-centric traffic information aggregating scheme. Integration of two types of traffic data sources, namely heterogeneous external real-time traffic information data sources and internal historical traffic information database, allowed the real-time traffic status to be envisaged through a scheme of knowledge-based system. Jiancheng et al. [29] determined the data processing flow through analyzing the charge transaction data of the electronic toll collection (ETC) data of Beijing City. Data processing techniques, together with index extraction techniques, were employed to extract the information supporting freeway operation and management. Brett and Roe [8] examined the clustering potential of the maritime transport sector in the Greater Dublin Region. Trullols et al. [53] developed a well-planned roadside infrastructure to distribute information in large-scale ITSs.

In 2011, Zhang et al. [66] conducted a survey on the development of the data-driven ITS (D2ITS), along with the functionality of all components and deployment issues. Jiang et al. [30] developed a model called the dynamic traffic assignment, for a continuum transportation system, to examine the characteristic traffic flow and the corresponding route choice behavior of travelers. Agamennoni et al. [1] deployed an unsupervised segmentation approach to sample the roads for inferring the topology of the network. The processing of logs of position traces to yield a representation of the road network has been accomplished with machine learning approaches. Lee et al. [37] developed a three-phase spatiotemporal traffic bottleneck mining (STBM) model involving numerous spatiotemporal traffic patterns, in addition to the STBM algorithms to discover the traffic bottlenecks associated with the urban network. Huang et al. [25] proposed a real-time intelligent statistical method to discriminate the driving condition of hybrid electric vehicles automatically.

Karlaftis and Vlahogianni [32] used neural networks and statistical methods to study the similarities and differences in transportation research. Faouzi et al. [15] investigated the use of data fusion in ITSs.

In 2012, Bhuyan and Rao [7] classified urban streets to form a wide range of classes, in addition to defining the ranges of speed concerned with the service level categories in Indian perspective. Zhu et al. [69] calibrated the traffic simulation model based on the data mining technique, which yields the vehicle’s speed. A new weighted regression method relying on agglomerative hierarchical cluster was also proposed. Wolfson et al. [59] developed a transportation query language (TranQuyl) for managing big data in ITSs. Jia and Yindong [28] scheduled and simulated the trip times based on automatic vehicle location (AVL) data. Caicedo et al. [10] predicted the parking space availability in real-time intelligent parking reservation systems using three subroutines. Fries et al. [16] examined the privacy challenges that are to be tackled while advancing the ITSs.

In 2013, Kompil and Celik [33] developed a trip distribution model. Kumar et al. [34] predicted the short-term traffic flow in heterogeneous conditions using artificial neural network and historical data from the Internet. Roman et al. [51] followed the service oriented architecture (SOA) and reference model of open distributed processing (RM-ODP) principles to design and develop an ITS architecture. Hu et al. [23] put forth a bi-level programming model of passenger transport network capacity to maximize the road network capacity. Yuwei and Pengfei [65] modified the neo-classical model that describes the regional economic growth in urban clusters using new economic geography theory and the tools aiding network testing. Weber et al. [57] suggested a novel system to analyze the conditions, determinants, and instruments that manage ServPPINs.

In 2014, Arabzad et al. [3] solved the multi-objective location-inventory problem concerned with a distribution network involving various third-party logistics (3PL) providers and transportation modes using an evolutionary algorithm. Zhang et al. [67] developed a hierarchical fuzzy rule-based system, which the genetic algorithms have optimized for envisaging the traffic congestion. He et al. [21] developed a novel vehicular data cloud service that is recommended for an Internet-of-Things (IoT) environment. An intelligent parking cloud service, along with the vehicular data mining cloud service, allowed the examination of vehicle warranty within an IoT environment. Ona et al. [47] diminished the heterogeneous opinions of public transport passengers through the clustering approach and service quality evaluation. Nahar and Sultana [44] developed a dynamic model that allows predicting the travel time, in case of the road networks, using artificial neural networks. Hsu et al. [22] developed a sequential pattern mining model-assisted eco-driving and cloud-based framework to lower the discharge of carbon dioxide and the intake of fuel in a sustainable ITS. Park et al. [48] used data mining strategies for real-time control of traffic in New York City. Li and Chen [38] envisaged travel time in a freeway, where non-recurrent congestion occurs. Arokhlo et al. [4] developed a new model for route planning that relies on Q-value-dependent dynamic programming to route vehicles among Malaysian cities.

Asif et al. [5] studied the spatiotemporal patterns, which have a lead role in the learning approaches that aid in predicting the speed of large-scale traffic. Large-scale prediction of road network and multiple prediction horizons with a support vector regression-based algorithm was also studied. He et al. [20] proposed an intelligent carpool routing scheme that mines the GPS trajectories to enable urban ride sharing. The eventual desire was to curtail the riding distance, lessen the expense associated with transportation, and to be free from urban traffic jams. Zheng and Wang [68] put forth an optimized scheme of frequent sets calculation, where the parallel NEclat and the cloud programming model work in unison. Malecki et al. [42] examined ITSs to reduce the environmental negative impact of urban freight transport in Szczecin. Monteiro et al. [43] structured an agent-based model to analyze the multimodal transportation network. Kannan et al. [31] developed the artificial immune system and sheep flock algorithms to produce solutions, whenever the two-stage fixed-charge transportation problem arises. Guo and Hall [18] designed a methodology for national freight origin-destination data to model the transportation in continental United States. Yangfan et al. [63] evaluated the traffic congestion mitigation in Beijing with variable message signs.

In 2015, Nezerenko et al. [46] organized the transportation of the Baltic Sea region using the cluster approach. Wibisono et al. [58] deployed the Fast Incremental Model Trees-Drift Detection (FIMT-DD) for the purpose of predicting the traffic big data visualization. Zonga and Wang [70] inspected three intertwining parking decisions, namely parking period, parking location, and parking duration, through the use of a Bayesian network. Wang et al. [56] predicted the personalized routes through employing the machine learning techniques to construct an effective probability transition matrix from big data. Two data reduction algorithms were also developed. Shi and Aty [52] examined the viability of a proactive real-time traffic monitoring strategy to assess its operation and safety. Pereira et al. [50] predicted public transport arrivals under special vents scenarios using web data. Pei et al. [49] proposed a density-based method to spot the two-component clusters. Necula [45] used R software (R & R of the Statistics Department of the University of Auckland, Auckland, New Zealand) with a set of libraries to analyze the traffic patterns on street segments with the extracted GPS data. Li et al. [39] examined data mining in a big data network, as well as the associated applications to predict the flow of traffic. Robust causal dependence mining of data was done. To carry out the regression as well as the causality analysis of the big data, a multiple-step strategy was introduced.

3 Trip Planning Using Data Mining Approaches

3.1 Descriptive Mining for Trip Planning

Ona et al. [47] employed latent class clustering to find the group of passengers with common perceptions about the quality of service, and discovered the variables that the passengers use to assess service quality. Zonga and Wang [70] developed a Bayesian network to examine parking choices like parking location, parking duration, and parking period, in addition to their effects, to develop measures that regulate the parking behavior in Beijing. Zheng and Wang [68] examined road transport management based on pruning Eclat algorithm, and MapReduce was used to analyze the demerits of the customary association rule mining approach. NEclat combined with cloud programming performed optimization to compute the measures.

Lee et al. [36] researched the generation of real-time information system, which relies on the user-centric traffic event reacting technique as well as the automatic agent-centric traffic information aggregating approach, and shared the framework that integrates the external and internal real-time dates for the ITS. The traffic information system method generated the traffic information. Jiancheng et al. [29] analyzed the charge transaction data of electronic toll collection system, which belong to Beijing City, to develop the model for travel speed computation in a freeway. By using the statistical analysis approach, they developed a travel speed calculation model. Kompil and Celik [33] found the latent abilities that the fuzzy and genetic fuzzy system approaches impart in modeling the urban trip distribution of Istanbul, and designed a new genetic fuzzy rule-based system and simple fuzzy rule-based system through the application of the Levenberg-Karguardt learning algorithm. He et al. [20] reduced the riding distances, transportation costs, and traffic jams in Beijing by developing a performance metric that is based on mining GPS data from shared riders of carpools, using a route-mining algorithm, rider selection, and route merging. Necula [45] identified the traffic patterns on the GPS-enabled data using R software, and determined whether the data are sufficient to draw inference about the traffic patterns. They applied K-means clustering, temporal mining, and spatial mining to predict the traffic pattern.

3.2 Predictive Mining for Trip Planning

Jiancheng et al. [29] analyzed the charge transaction data, which are concerned with the electronic toll collection system in Beijing City, to create the model for travel speed computation in a freeway using the statistical analysis approach. Agamennoni et al. [1] developed an unsupervised segmentation approach that enables roads to be sampled to infer the network topology. The road network details were obtained from the vehicle’s path, and hidden Markov model produced the transition matrix. Jiang et al. [30] developed a predictive continuum dynamic user-optimal model and inspected the dynamic flow characteristics and the route choice of travelers in an urban road. Finite volume, finite element, and the Runge-Kutta time stepping method were used to develop the model. Lee et al. [37] found the traffic bottlenecks by using the spatiotemporal mining method and implemented this system in relation to the taxi dispatching system of urban networks like Taipei and Taiwan. The spatial heuristic clustering method aided the traffic bottleneck identification. Kumar et al. [34] predicted the short-term traffic flow in heterogeneous conditions using the artificial neural network, and analyzed the sensitivity of the best neural network. Artificial neural network simulated the system’s structural and functional aspects. The relationship between the input and the output variables was obtained using sensitivity analysis. For evaluation of the predicted model, they used the coefficient of correlation and root mean square error measures.

Arabzad et al. [3] developed an optimization algorithm with multiple objectives to solve the multi-objective location inventory problem, in which a distribution center network that is constituted of various third-party logistic providers as well as transportation mode suffers. The non-dominated sorting genetic model was applied to perform high-quality searching. Evaluation was done using quality metric, spacing metric, and diversity metric. The developed algorithm promoted the decision-making process. Zhang et al. [67] developed a hierarchical fuzzy rule-based system, in which the genetic algorithm performed optimization, to predict a system using big data to yield traffic congestion information with high robustness and accuracy. In addition, the outcomes from the representative algorithm were also compared with those obtained with it. The fuzzy rule-based system was developed using fuzzy AdaBoost, fuzzy LogitBoost, fuzzy Chi-RW, and GHFRBS. Nahar and Sultana [44] developed a dynamic model for estimating the time of travel to support the road networks. They compared their method with a link-dependent prediction model and time-varying coefficient linear regression model using the magnitude of the relative error. The results showed that the artificial neural network serves better in analyzing and predicting the traffic data.

Asif et al. [5] developed a model that predicts large interconnected road networks by using a support vector regression-based algorithm, and studied the effectiveness of the developed performance analysis methods. The support vector regression method predicted traffic with high accuracy, and mean absolute percentage error was the used prediction error measure. Wibisono et al. [58] predicted the traffic condition in UK road transportation through the FIMT-DD method, and visualized the traffic flow predicted model within the sensor point that was generated after the real map simulation. The FIMT-DD algorithm functioned perfectly in all the incremental learning of regression fields, and the traffic prediction error was obtained using the mean absolute error. Wang et al. [56] suggested a machine learning algorithm for modeling route prediction using a Markov chain model. Additionally, an algorithm that relies on the probability transition matrix to produce two data reduction algorithms was presented, and its aim was to predict the routes. The two data reduction processes serve two purposes: (i) to perform a mapping between the massive GPS data and a compact link-based system and (ii) to diminish the size concerned with a probability transition matrix. Li et al. [39] worked on the issue of predicting traffic with improved prominence given to big data handling for developing a Lasso Granger causality theory. The Lasso method was used to obtain the desired information, omitting the unwanted data. All the measurements were obtained using Gaussian distribution and mean square deviation.

4 Clustering Models for Transportation Data Analysis

4.1 Unsupervised Models

Bhuyan and Rao [7] created a model to estimate the travel speed in urban streets, which aids in grouping the roads and defining the level of service, with K-means as well as K-medoid. Metrics like separation index, Dunn’s index, Xie and Ben’s index, and partition index determined the precision of the estimated traveling speed. Their study revealed that the K-medoid clustering method offers better prediction results. Zhu et al. [69] developed a method involving agglomerative hierarchical clustering to calibrate the parameters of the speed-density traffic simulation model. Other techniques like K-means were also used by them to calibrate the speed-density parameter, and their performance index was measured as root mean square percent error. Nezerenko et al. [46] developed a model that achieves new quantitative transportation field in the Baltic Sea region by forming a single regional transportation. Hierarchical clustering analysis, Bayesian analysis, affinity analysis, and correlation analysis were used to identify countries with homogenous trends in the transportation field. Pei et al. [49] proposed a method that creates a two-component cluster to spot the clusters of taxi-cab trips in Beijing City (Figure 2).

Figure 2: List of Unsupervised Models Used in Intelligent Transportation System.

Figure 2:

List of Unsupervised Models Used in Intelligent Transportation System.

4.2 Supervised Models

Jacob and Abdulhai [26] developed a self-learning adaptive integrated freeway arterial control, in order to support the handling of non-recurring as well as recurring congestion in North America. To predict the model, they used algorithms like back-propagation, K-nearest neighbor, and Q-learning. Li and Chen [38] predicted the travel time of a freeway with non-recurrent congestion using K-means clustering, decision trees, and neural network. Arokhlo et al. [4] constructed a model for a route planning system, which is based on multi-agent reinforcement learning algorithm, for the purpose of routing vehicles between Malaysian cities. They studied the components, constituting the road network areas, and their associated weights for solving the vehicle delay problems using the Q-value-based dynamic programming and Boltzmann distribution. The average traveling time was computed to find the best route for vehicles to avoid delays. Pereira et al. [50] predicted the public transport arrival in event areas by using off-the-shelf techniques for data collection from Singapore City. To predict, they used various machine learning algorithms, such as K-nearest neighbor, Gaussian process, linear regression, and support vector regression. Zhu et al. [69] suggested methods to calibrate the parameters of the speed-density traffic simulation model to calculate the speed of vehicles. They used K-nearest neighbor and locally weighted regression for the speed-density parameter calibration, and the performance index used here was the measured root mean square percent error (Figure 3).

Figure 3: List of Supervised Models Used in Intelligent Transportation System.

Figure 3:

List of Supervised Models Used in Intelligent Transportation System.

5 Findings and Discussion

5.1 State-of-the-Art Reviews

In ITS, cluster analysis has wide usage. Clustering experiments have classified big data and defined the speed ranges for the service-level categories. It has been the only suitable solution technique for various classification problems. Bhuyan and Rao [7] randomly opted a few groups from a substantially large observation. Zhu et al. [69] yielded the highest precision and accurate traffic dynamics by using it, promoting effective data processing in large-scale road networks. Pei et al. [49] clustered data with two types of points in ITS. Nezerenko et al. [46] identified more homogeneous groups of users and formed a single transportation system using clustering. Ona et al. [47] reduced the heterogeneous opinions of service quality public transport passengers using cluster analysis (Figures 4 and 5).

Figure 4: List of Pattern Mining Methods Used in Intelligent Transportation System.

Figure 4:

List of Pattern Mining Methods Used in Intelligent Transportation System.

Figure 5: List of Predictive Methods Used in Intelligent Transportation System.

Figure 5:

List of Predictive Methods Used in Intelligent Transportation System.

5.2 Research Gaps

Unsupervised clustering models are widely used; however, a few drawbacks still exist, as follows: (i) Determining the optimal number of clustering points seems to be a major problem while developing a model that is based on clustering methods. (ii) Defining a termination criterion is difficult in case of unsupervised clustering models. Clustering a specific type of data and finding its termination criterion seems to be ineffective because most unsupervised algorithms are iterative. If the criterion is met for all clustered data, the developed model can be trained robustly. (iii) Clustering methods are sensitive that even unwanted data may be amplified and interpreted. Thus, there is a lack of a definitive objective function.

Evolutionary computational algorithm overcomes the aforesaid problems. It is robust and suited well to high-dimensional problems. Yet, these techniques also have a few drawbacks, as follows: (i) Accomplishing reliable results within the stipulated period is challenging. (ii) Adverse configurations lead to premature convergence. Hence, the population diversity gets lost rapidly to a local extremum. (iii) Anytime behavior is rather common in the evolutionary approach. The algorithm shows rapid progress at the initial stage and flatters anytime.

6 Conclusion and Future Work

This paper has reviewed various data mining methods for ITS. Descriptive mining methods have aided in generating a real-time information system, identifying traffic patterns, developing travel speed calculation model, and investigating parking decisions. Predictive data mining techniques have helped in inferring the network topology, finding traffic bottlenecks, solving the multi-objective location inventory problem, constructing two data reduction algorithms, and predicting short-term traffic flow in heterogeneous conditions. Among the descriptive data mining methods, clustering models have extensive use in forming the single transportation system, calibrating the speed-density parameters, estimating the travel speed, and clustering the taxi-cab trips and route planning system. Especially, the unsupervised methods have many applications in ITS. Traditional unsupervised models have drawbacks; however, these are rectified using the evolutionary approach. The presented evolutionary approach also has drawbacks. Thus, increasing the potential of the evolutionary model is rather challenging. An effective evolutionary approach without drawbacks will certainly help in developing enhanced ITS.


[1] G. Agamennoni, J. I. Nieto and E. M. Nebot, Robust inference of principal road paths for Intelligent Transportation Systems, IEEE Trans. Intell. Transport. Syst.12 (2011), 298–308.10.1109/TITS.2010.2069097Search in Google Scholar

[2] W. Andreas, W. Shangbo, B. Guido and J. Peter, Traffic congestion estimation service exploiting mobile assisted positioning schemes in GSM networks, Proc. Earth Planet. Sci.1 (2009), 1385–1392.10.1016/j.proeps.2009.09.214Search in Google Scholar

[3] S. M. Arabzad, M. Ghorbani and R. T. Moghaddam, An evolutionary algorithm for a new multi-objective location-inventory model in a distribution network with transportation modes and third-party logistics providers, Int. J. Prod. Res.53 (2014), 1038–1050.10.1080/00207543.2014.938836Search in Google Scholar

[4] M. Z. Arokhlo, A. Selamat, S. Z. M. Hashim and H. Afkhami, Modeling of route planning system based on Q value-based dynamic programming with multi-agent reinforcement learning algorithms, Eng. Appl. Artif. Intell.29 (2014), 163–177.10.1016/j.engappai.2014.01.001Search in Google Scholar

[5] M. T. Asif, J. Dauwels, C. Y. Goh, A. Oran, E. Fathi, M. Xu, M. M. Dhanya, N. Mitrovic and P. Jaillet, Spatiotemporal patterns in large-scale traffic speed prediction, IEEE Trans. Intell. Transport. Syst.15 (2014), 794–804.10.1109/TITS.2013.2290285Search in Google Scholar

[6] S. Bamberg, S. Fujii, M. Friman and T. Garling, Behaviour theory and soft transport policy measures, Transport Policy18 (2011), 228–235.10.1016/j.tranpol.2010.08.006Search in Google Scholar

[7] P. K. Bhuyan and K. V. K. Rao, Defining los criteria of urban streets using GPS data: k-means and k-medoid clustering in Indian context, Transport27 (2011), 149–157.10.3846/16484142.2012.692354Search in Google Scholar

[8] V. Brett and M. Roe, The potential for the clustering of the maritime transport sector in the Greater Dublin Region, Marit. Pol. Mgmt.37 (2010), 1–16.10.1080/03088830903461126Search in Google Scholar

[9] Y. Byon, B. Abdulhai and A. Shalaby, Real-time transportation mode detection via tracking Global Positioning System mobile devices, J. Intell. Transport. Syst.13 (2009), 161–170.10.1080/15472450903287781Search in Google Scholar

[10] F. Caicedo, C. Blazquez and P. Miranda, Prediction of parking space availability in real time, Expert Syst. Appl.39 (2012), 7281–7290.10.1016/j.eswa.2012.01.091Search in Google Scholar

[11] S. W. Chiou, A cutting plane projection method for bi-level area traffic control optimization with uncertain travel demand, Appl. Math. Comput.266 (2015), 390–403.10.1016/j.amc.2015.05.009Search in Google Scholar

[12] T. G. Crainic, M. Gendreau and J. Y. Potvin, Intelligent freight-transportation systems: assessment and the contribution of operations research, Transport. Res.17 (2009), 541–557.10.1016/j.trc.2008.07.002Search in Google Scholar

[13] W. Davidson, R. Donnelly, P. Vovsha, J. Freedman, S. Ruegg, J. Hicks, J. Castiglione and R. Picado, Synthesis of first practices and operational research approaches in activity-based travel demand modeling, Transport. Res. Pt. A Policy Pract.41 (2007), 464–488.10.1016/j.tra.2006.09.003Search in Google Scholar

[14] Y. Duvarci and S. Mizokami, A suppressed demand analysis method of the transportation disadvantaged in policy making, Transport. Plann. Technol.32 (2009), 187–214.10.1080/03081060902861400Search in Google Scholar

[15] N. E. E. Faouzi, H. Leung and A. Kurian, Data fusion in intelligent transportation systems: progress and challenges – a survey, Inform. Fusion12 (2011), 4–10.10.1016/j.inffus.2010.06.001Search in Google Scholar

[16] R. N. Fries, M. R. Gahrooei, M. Chowdhury and A. J. Conway, Meeting privacy challenges while advancing intelligent transportation systems, Transport. Res. Pt. C25 (2012), 34–45.10.1016/j.trc.2012.04.002Search in Google Scholar

[17] M. Friman, L. Larhult and T. Garling, An analysis of soft transport policy measures implemented in Sweden to reduce private car use, Transportation40 (2013), 109–129.10.1007/s11116-012-9412-ySearch in Google Scholar

[18] F. Guo and L. A. Hall, A zone design methodology for national freight origin – destination data and transportation modeling, Transport. Plann. Technol.37 (2014), 738–756.10.1080/03081060.2014.959355Search in Google Scholar

[19] M. Hatzopoulou and E. J. Miller, Linking an activity-based travel demand model with traffic emission and dispersion models: transport’s contribution to air pollution in Toronto, Transport. Res. Pt. D Transport Environ.15 (2010), 315–325.10.1016/j.trd.2010.03.007Search in Google Scholar

[20] W. He, K. Hwang and D. Li, Intelligent carpool routing for urban ridesharing by mining GPS trajectories, IEEE Trans. Intell. Transport. Syst.15 (2014), 2286–2296.10.1109/TITS.2014.2315521Search in Google Scholar

[21] W. He, G. Yan and L. D. Xu, Developing vehicular data cloud services in the IoT environment, IEEE Trans. Indust. Inform.10 (2014), 1587–1595.10.1109/TII.2014.2299233Search in Google Scholar

[22] C. Y. Hsu, C. S. Yang, L. C. Yu, C. F. Lin, H. H. Yao, D. Y. Chen, K. R. Lai and P. C. Chang, Development of a cloud-based service framework for energy conservation in a sustainable intelligent transportation system, Int. J. Prod. Econ.164 (2014), 454–461.10.1016/j.ijpe.2014.08.014Search in Google Scholar

[23] Y. Hu, Z. Xiang and H. Chen, Selection and peer-review under responsibility of Chinese Overseas Transportation Association (COTA), Proc. Soc. Behav. Sci.96 (2013), 1861–1867.10.1016/j.sbspro.2013.08.211Search in Google Scholar

[24] B. Huang, Q. Wu and F. B. Zhan, A shortest path algorithm with novel heuristics for dynamic transportation networks, Int. J. Geogr. Inform. Sci.21 (2007), 625–644.10.1080/13658810601079759Search in Google Scholar

[25] X. Huang, Y. Tan and X. He, An intelligent multi feature statistical approach for the discrimination of driving conditions of a hybrid electric vehicle, IEEE Trans. Intell. Transport. Syst.12 (2011), 453.10.1109/TITS.2010.2093129Search in Google Scholar

[26] C. Jacob and B. Abdulhai, Machine learning for multi-jurisdictional optimal traffic corridor control, Transport. Res.44 (2010), 53–64.10.1016/j.tra.2009.11.001Search in Google Scholar

[27] A. Jarasuniene, Research into intelligent transport systems, Transport22 (2007), 61–67.10.3846/16484142.2007.9638100Search in Google Scholar

[28] X. Jia and S. Yindong, Setting scheduled trip time based on AVL data, J. Transport. Syst. Eng. Inform. Technol.12 (2012), 39–45.10.1016/S1570-6672(11)60222-5Search in Google Scholar

[29] J. Jiancheng, L. Lili and D. Bo, ETC data based traffic information mining techniques, J. Transport. Syst. Eng. Inform. Technol.10 (2010), 57–63.10.1016/S1570-6672(09)60033-7Search in Google Scholar

[30] Y. Jiang, S. C. Wong, H. W. Ho, P. Zhang, R. Liu and A. Sumalee, A dynamic traffic assignment model for a continuum transportation system, Transport. Res. Pt. B45 (2011), 343–363.10.1016/j.trb.2010.07.003Search in Google Scholar

[31] D. Kannan, K. Govindan and H. Soleimanic, Artificial immune system and sheep flock algorithms for two-stage fixed-charge transportation problem, Optimization63 (2014), 1465–1479.10.1080/02331934.2014.898148Search in Google Scholar

[32] M. G. Karlaftis and E. I. Vlahogianni, Statistical methods versus neural networks in transportation research: differences, similarities and some insights, Transport. Res. Pt. C19 (2011), 387–399.10.1016/j.trc.2010.10.004Search in Google Scholar

[33] M. Kompil and H. M. Celik, Modelling trip distribution with fuzzy and genetic fuzzy systems, Transport. Plann. Technol.36 (2013), 170–200.10.1080/03081060.2013.770946Search in Google Scholar

[34] K. Kumar, M. Parida and V. K. Katiyar, Short term traffic flow prediction in heterogeneous condition using artificial neural network, Transport. Plann. Technol.36 (2013), 170–200.10.3846/16484142.2013.818057Search in Google Scholar

[35] W. H. Lee, S. S. Tseng and S. H. Tsai, A knowledge based real-time travel time prediction system for urban network, Expert Syst. Appl.36 (2009), 4239–4247.10.1016/j.eswa.2008.03.018Search in Google Scholar

[36] W. H. Lee, S. S. Tseng and W. Y. Shieh, Collaborative real-time traffic information generation and sharing framework for the intelligent transportation system, Inform. Sci.180 (2010), 62–70.10.1016/j.ins.2009.09.004Search in Google Scholar

[37] W. H. Lee, S. S. Tseng, J. L. Shieh and H. H. Chen, Discovering traffic bottlenecks in an urban network by spatiotemporal datamining on location-based services, IEEE Trans. Intell. Transport. Syst.12 (2011), 1047–1056.10.1109/TITS.2011.2144586Search in Google Scholar

[38] C. S. Li and M. C. Chen, A datamining based approach for travel time prediction in freeway with non-recurrent congestion, Neurocomputing133 (2014), 74–83.10.1016/j.neucom.2013.11.029Search in Google Scholar

[39] L. Li, X. Su, Y. Wanga, Y. Lin, Z. Li and Y. Li, Robust causal dependence mining in big data network and its application to traffic flow predictions, Transport. Res. Pt. C58 (2015), 292–307.10.1016/j.trc.2015.03.003Search in Google Scholar

[40] Y. Liang, M. L. Reyes and J. D. Lee, Real-time detection of driver cognitive distraction using support vector machines, IEEE Trans. Intell. Transport. Syst.8 (2007), 340–350.10.1109/TITS.2007.895298Search in Google Scholar

[41] H. Z. Lin, H. P. Lo and X. J. Chen, Lifestyle classifications with and without activity-travel patterns, Transport. Res. Pt. A Policy Pract.43 (2009), 626–638.10.1016/j.tra.2009.04.002Search in Google Scholar

[42] K. Malecki, S. Iwan and K. Kijewska, Influence of intelligent transportation systems on reduction of the environmental negative impact of urban freight transport based on Szczecin example, Proc. Soc. Behav. Sci.151 (2014), 215–229.10.1016/j.sbspro.2014.10.021Search in Google Scholar

[43] N. Monteiro, R. Rossetti, P. Campos and Z. Kokkinogenis, A framework for a multimodal transportation network: an agent-based model approach, Transport. Res. Proc.4 (2014), 213–227.10.1016/j.trpro.2014.11.017Search in Google Scholar

[44] L. Nahar and Z. Sultana, A new travel time prediction method for intelligent transportation system, IOSR J. Comput. Eng.16 (2014), 24–30.10.9790/0661-16382430Search in Google Scholar

[45] E. Necula, Analyzing traffic patterns on street segments based on GPS data using R, Transport. Res. Proc.10 (2015), 276–285.10.1016/j.trpro.2015.09.077Search in Google Scholar

[46] O. Nezerenko, O. Koppe and T. Tuisk, Cluster approach in organization of transportation in the Baltic Sea region, Transport30 (2015), 1–13.10.3846/16484142.2014.994225Search in Google Scholar

[47] R. D. Ona, G. Lopez, F. J. D. Rios and J. D. Ona, Cluster analysis for diminishing heterogeneous opinions of service quality public transport passengers, Proc. Soc. Behav. Sci.162 (2014), 459–466.10.1016/j.sbspro.2014.12.227Search in Google Scholar

[48] H. J. Park, P. H. Kim, M. Marsico and N. Rasheed, Datamining strategies for real-time control in New York City, Proc. Comput. Sci.32 (2014), 109–116.10.1016/j.procs.2014.05.404Search in Google Scholar

[49] T. Pei, W. Wang, H. Zhang, T. Ma, Y. Du and C. Zhou, Density-based clustering for data containing two types of points, Int. J. Geogr. Inform. Sci.29 (2015), 175–193.10.1080/13658816.2014.955027Search in Google Scholar

[50] F. C. Pereira, F. Rodrigues and M. B. Akiva, Using data from the web to predict public transport arrivals under special events scenarios, J. Intell. Transport. Syst.19 (2015), 273–288.10.1080/15472450.2013.868284Search in Google Scholar

[51] I. Roman, G. Madinabeitia, L. Jimenez, G. A. Molina and J. A. Ternero, Experiences applying RM-ODP principles and techniques to intelligent transportation system architectures, Comput. Stand. Interf.35 (2013), 338–347.10.1016/j.csi.2011.12.004Search in Google Scholar

[52] Q. Shi and M. A. Aty, Big data applications in real-time traffic operation and safety monitoring and improvement on urban expressways, Transport. Res. Pt. C58 (2015), 380–394.10.1016/j.trc.2015.02.022Search in Google Scholar

[53] O. Trullols, M. Fiore, C. Casetti, C. F. Chiasserini and J. M. B. Ordinas, Planning roadside infrastructure for information dissemination in intelligent transportation systems, Comput. Commun.33 (2010), 432–442.10.1016/j.comcom.2009.11.021Search in Google Scholar

[54] A. Vij and K. Shankari, When is big data big enough? Implications of using GPS-based surveys for travel demand analysis, Transport. Res. Pt. C Emerg. Technol.56 (2015), 446–462.10.1016/j.trc.2015.04.025Search in Google Scholar

[55] C. Wang, M. Quddus, M. Enoch, T. Ryley and L. Davison, Exploring the propensity to travel by demand responsive transport in the rural area of Lincolnshire in England, Case Stud. Transport Policy3 (2015), 129–136.10.1016/j.cstp.2014.12.006Search in Google Scholar

[56] X. Wang, Y. Ma, J. Di, Y.L. Murphey, S. Qiu, J. Kristinsson, J. Meyer, F. Tseng and T. Feldkamp, Building efficient probability transition matrix using machine learning from big data for personalized route prediction, Proc. Comput. Sci.53 (2015), 284–291.10.1016/j.procs.2015.07.305Search in Google Scholar

[57] K. M. Weber, B. H. Schuh, H. Godoe and R. Roeste, ICT-enabled system innovations in public services: experiences from intelligent transport systems, Telecommun. Policy38 (2013), 539–557.10.1016/j.telpol.2013.12.004Search in Google Scholar

[58] A. Wibisono, W. Jatmiko, H. A. Wisesa, B. Hardjono and P. Mursanto, Traffic big data prediction and visualization using Fast Incremental Model Trees-Drift Detection (FIMT-DD), Knowl. Based Syst.93 (2016), 33–46.10.1016/j.knosys.2015.10.028Search in Google Scholar

[59] O. Wolfson, A. P. Sistla and B. Xu, The TranQuyl language for data management in intelligent transportation, Transport. Res. Pt. C23 (2012), 3–13.10.1016/j.trc.2012.02.002Search in Google Scholar

[60] J. Xia and M. Chen, Defining traffic flow phases using Intelligent Transportation Systems-generated data, J. Intell. Transport. Syst.11 (2007), 15–24.10.1080/15472450601122322Search in Google Scholar

[61] P. Yadav, Document-document similarity matrix and multiple-kernel fuzzy C-means algorithm based web document clustering for information retrieval, Int. J. Adv. Res. Comput. Commun. Eng.3 (2014), 8317–8321.10.17148/IJARCCE.2014.31054Search in Google Scholar

[62] P. Yadav, SR-K-Means clustering algorithm for semantic information retrieval, Int. J. Invent. Comput. Sci. Eng.1 (2014), 17–24.Search in Google Scholar

[63] Z. Yangfan, J. Shunping, G. Wei and L. Shuang, Evaluation on traffic congestion mitigation in Beijing with variable message signs, J. Transport. Syst. Eng. Inform. Technol.14 (2014), 201–206.10.1016/S1570-6672(13)60146-4Search in Google Scholar

[64] Y. Yin, S. M. Madanat and X. Y. Lu, Robust improvement schemes for road networks under demand uncertainty, Eur. J. Oper. Res.198 (2009), 470–479.10.1016/j.ejor.2008.09.008Search in Google Scholar

[65] L. Yuwei and N. Pengfei, Externalities, transport networks and the economic growth of urban clusters, Special issue: regional convergence and coordinated urban growth, Soc. Sci. China34 (2013), 174–194.10.1080/02529203.2013.820565Search in Google Scholar

[66] J. Zhang, F. Y. Wang, K. Wang, W. H. Lin, X. Xu and C. Chen, Data-driven intelligent transportation systems: a survey, IEEE Trans. Intell. Transport. Syst.12 (2011), 1624–1639.10.1109/TITS.2011.2158001Search in Google Scholar

[67] X. Zhang, E. Onieva, A. Perallos, E. Osaba and V. C. S. Lee, Hierarchical fuzzy rule-based system optimized with genetic algorithms for short term traffic congestion prediction, Transport. Res. Pt. C43 (2014), 127–142.10.1016/j.trc.2014.02.013Search in Google Scholar

[68] X. Zheng and S. Wang, Study on the method of road transport management information datamining based on pruning Eclat Algorithm and MapReduce, Proc. Soc. Behav. Sci.138 (2014), 757–766.10.1016/j.sbspro.2014.07.254Search in Google Scholar

[69] J. Zhu, L. Shubin and L. Xiaoqing, Parameters calibration of traffic simulation model based on datamining, J. Transport. Syst. Eng. Inform. Technol.12 (2012), 28–33.10.1016/S1570-6672(11)60232-8Search in Google Scholar

[70] F. Zonga and M. Wang, Understanding parking decisions with a Bayesian network, Transport. Plann. Technol.38 (2015), 585–600.10.1080/03081060.2015.1048943Search in Google Scholar

Received: 2016-6-15
Published Online: 2017-3-7
Published in Print: 2018-3-28

©2018 Walter de Gruyter GmbH, Berlin/Boston

This article is distributed under the terms of the Creative Commons Attribution Non-Commercial License, which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.