Hyunjo Lee

A New Travel Time Prediction Method for Intelligent Transportation Systems

Travel time prediction is an indispensable for numerous intelligent transportation systems (ITS) including advanced traveler information systems. The main purpose of this research is to develop a dynamic travel time prediction model for road networks. In this paper we propose a new method to predict travel times using Naive Bayesian Classification (NBC) model because Naive Bayesian Classification has exhibited high accuracy and speed when applied to large databases. Our proposed prediction algorithm is also scalable to road networks with arbitrary travel routes. In addition, we compare the proposed method with such prediction methods as link-based prediction model and time-varying coefficient linear regression model. It is shown from our experiment that NBC predictor can reduce mean absolute relative error significantly rather than the other predictors. We illustrate the practicability of applying NBC in travel time prediction and prove that NBC is suitable and performs well for traffic data analysis.

New travel time prediction algorithms for intelligent transportation systems

Recently, travel time prediction has become a crucial part of trip panning and dynamic route guidance for many advanced traveler information and transportation management systems. Moreover, a scalable prediction system with high accuracy is critical for the successful deployment of ATIS (Advanced Travelers Information Systems) in road networks. In this paper, we propose two travel time prediction algorithms using naive Bayesian classification and rule-based classification. Both classification techniques provide a velocity class to be used for measuring travel time accurately. Our algorithms exhibit high accuracy in predicting travel time when using a large amount of historical traffic database. In addition, our travel time prediction algorithms are suitable for road networks with arbitrary travel routes. It is shown from our performance comparison, our travel time prediction algorithms significantly outperform the existing prediction algorithms, such as the link-based algorithm, the switching model, and the linear regression algorithm. In addition, it is revealed that our algorithm using naive Bayesian classification is better on the performance of mean absolute relative error than our algorithm using rule-based classification.

Modified K-means clustering for travel time prediction based on historical traffic data

Prediction of travel time has major concern in the research domain of Intelligent Transportation Systems (ITS). Clustering strategy can be used as a powerful tool of discovering hidden knowledge that can easily be applied on historical traffic data to predict accurate travel time. In our Modified K-means Clustering (MKC) approach, a set of historical data is portioned into a group of meaningful sub-classes (also known as clusters) based on travel time, frequency of travel time and velocity for a specific road segment and time group. With the use of same set of historical travel time estimates, comparison is also made to the forecasting results of other three methods: Successive Moving Average (SMA), Chain Average (CA) and Naive Bayesian Classification (NBC) method. The results suggest that the travel times for the study periods could be predicted by the proposed method with the minimum Mean Absolute Relative Error (MARE).

Context-Aware Architecture for Intelligent Application Services in Ubiquitous Computing

Topic-based language model has attracted much attention as the propounding of semantic retrieval in recent years. Especially for the ASR text with errors, the topic representation is more reasonable than the exact term representation. Among these models, Latent Dirichlet Allocation(LDA) has been noted for its ability to discover the latent topic structure, and is broadly applied in many text-related tasks. But up to now its application in information retrieval(IR) is still limited to be a supplement to the standard document models, and furthermore, it has been pointed out that directly employing the basic LDA model will hurt retrieval performance. In this paper, we propose a lexicon-guided two-level LDA retrieval framework. It uses the HowNet to guide the first-level LDA models parameter estimation, and further construct the second-level LDA models based on the first-levels inference results. We use TRECID 2005 ASR collection to evaluate it, and compare it with the vector space model(VSM) and latent semantic Indexing(LSI). Our experiments show the proposed method is very competitive.In this paper, we design a context-aware architecture for dealing with intelligent application services in ubiquitous computing. The context-aware architecture is composed of middleware, context server, and client. The middleware component of our context-aware architecture plays an important role in recognizing a moving node with mobility by using a Bluetooth wireless communication technology as well as in executing an appropriate execution module according to the context acquired from a context server. The context server functions as a manager that efficiently stores into the database server context information, such as users current status, physical environment, and resources of a computing system. To verify the usefulness of our architecture, we finally develop a context-aware application system base on it, which provides users with a music playing service in ubiquitous computing environment.

A New Energy-Efficient Cluster-Based Routing Protocol Using a Representative Path in Wireless Sensor Networks

Wireless sensor networks (WSNs) have been broadly studied with advances in ubiquitous computing environment. Because the resource of a sensor node is limited, it is important to use energy-efficient routing protocol in WSNs. The cluster-based routing is an efficient way to reduce energy consumption by decreasing the number of transmitted messages to the sink node. LEACH is the most popular cluster-based routing protocol, which provides an adaptive cluster generation and cluster header rotation. However, its communication range is limited since it assumes a direct communication between sensor nodes and a sink node. To resolve this problem, we propose a new energy-efficient cluster-based routing protocol, which adopts a centralized clustering approach to select cluster headers by generating a representative path. To support reliable data communication, we propose a multihop routing protocol that allows both intra- and intercluster communications. Based on a message success rate and a representative path, the sensor nodes are uniformly distributed in clusters so that the lifetime of network can be prolonged. Through performance analysis, we show that our energy-efficient routing protocol outperforms the existing protocols up to 2 times, in terms of the distribution of cluster members, the energy consumption, and the reliability of a sensor network.

Development of an Effective Travel Time Prediction Method Using Modified Moving Average Approach

Prediction of travel time on road network has emerged as a crucial research issue in intelligent transportation system (ITS). Travel time prediction provides information that may allow travelers to change their routes as well as departure time. To provide accurate travel time for travelers is the key challenge in this research area. In this paper, we formulate two new methods which are based on moving average can deal with this kind of challenge. In conventional moving average approach, data may lose at the beginning and end of a series. It may sometimes generate cycles or other movements that are not present in the original data. Our proposed modified method can strongly tackle those kinds of uneven presence of extreme values. We compare the proposed methods with the existing prediction methods like Switching method [10] and NBC method [11]. It is also revealed that proposed methods can reduce error significantly in compared with other existing methods.

Hilbert-Curve Based Data Aggregation Scheme to Enforce Data Privacy and Data Integrity for Wireless Sensor Networks

Data aggregation techniques have been proposed for wireless sensor networks (WSNs) to address the problems presented by the limited resources of sensor nodes. The provision of efficient data aggregation to preserve data privacy is a challenging issue in WSNs. Some existing data aggregation methods for preserving data privacy are CPDA, SMART, the Twin-Key based method, and GP2S. These methods, however, have two limitations. First, the communication cost for network construction is considerably high. Second, they do not support data integrity. There are two methods for supporting data integrity, iCPDA and iPDA. But they have high communication cost due to additional integrity checking messages. To resolve this problem, we propose a novel Hilbert-curve based data aggregation scheme that enforces data privacy and data integrity for WSNs. To minimize communication cost, we utilize a tree-based network structure for constructing networks and aggregating data. To preserve data privacy, we make use of both a seed exchange algorithm and Hilbert-curve based data encryption. To support data integrity, we use an integrity checking algorithm based on the PIR technique by directly communicating between parent and child nodes. Finally, through a performance analysis, we show that our scheme outperforms the existing methods in terms of both energy efficiency and privacy preservation.

A new cloaking algorithm using Hilbert curves for privacy protection

Due to the advancement of GPS facilitates, the use of Location Based Service (LBS) has recently been increased rapidly. Since LBS needs the location of user, the private and confidential information of user may disclose to others. To protect the privacy of users, many cloaking algorithms have been proposed to hide users actual location. The existing Hilbert cloaking algorithm shows its high accuracy in terms of location privacy, but it has a drawback that it extends a cloaking region inefficiently due to the dimensionality reduction. In this paper, we propose a new cloaking algorithm which can avoid the unnecessary extension of cloaking region. Our algorithm optimizes the generation of a cloaking region by storing adjacent cell information being not connected by Hilbert curve. From experimental results, it is shown that our proposed cloaking algorithm outperforms the existing Hilbert algorithm.

A Travel Time Prediction Algorithm Using Rule-Based Classification on MapReduce

Recently, the amount of trajectory data has been rapidly increasing with the popularity of LBS and the development of mobile technology. Thus, the analysis of trajectory patterns for large amounts of trajectory data has attracted much interest. To improve the quality of trajectory-based services, it is essential to predict an exact travel time for a given query on road networks. One of the typical schemes for travel time prediction is a rule-based classification method which can ensure high accuracy. However, the existing scheme is inadequate for the processing of massive data because it is designed without the consideration of distributed computing environments. To solve this problem, this paper proposes a travel time prediction algorithm using rule-based classification on MapReduce for a large amount of trajectory data. First, our algorithm generates classification rules based on the actual traffic statistics and measures adequate velocity classes for each road segment. Second, our algorithm generates a distributed index by using the grid-based map partitioning method. Our algorithm can reduces the query processing cost because it only retrieves the grid cells which contain a query region, instead of the entire road network. Furthermore, it can reduce the query processing time by estimating the travel time for each segment of a given query in a parallel way. Finally, we show from our performance analysis that our scheme performs more accurate travel time prediction than the existing algorithms.

An Efficient High-Dimensional Indexing Scheme Using a Clustering Technique for Content-Based Retrieval

Since video data, like UCC(User Created Contents), has recently attracted much interest, high-dimensional indexing schemes are required to satisfy users’ requirements. However, except Hybrid Spill-Tree, the existing high-dimensional indexing schemes are not efficient in terms of retrieval performance because they are weak in either retrieval accuracy or retrieval time. Therefore, we, in this paper, propose a new efficient high-dimensional indexing scheme to satisfy users’ requirements by supporting the content-based retrieval of a large amount of video data. For this, we extend Hybrid Spill-Tree with a signature-based clustering technique. In addition, we provide both an insertion algorithm and a k-NN search algorithm for our high-dimensional indexing scheme. Finally, we show that our high-dimensional indexing scheme achieves better retrieval performance than M-Tree and Hybrid Spill-Tree.