We propose a new dynamic method for multidimensional selectivity estimation for range queries tha... more We propose a new dynamic method for multidimensional selectivity estimation for range queries that works accurately independent of data distribution. Good estimation of selectivity is important for query optimization and physical database design. Our method employs the multilevel grid file (MLGF) for accurate estimation of multidimensional data distribution. The MLGF is a dynamic, hierarchical, balanced, multidimensional file structure that gracefully adapts to nonuniform and correlated distributions. We show that the MLGF directory naturally represents a multidimensional data distribution. We then extend it for further refinement and present the selectivity estimation method based on the MLGE Extensive experiments have been performed to test the accuracy of selectivity estimation. The results show that estimation errors are very small independent of distributions, even with correlated and/or highly skewed ones. Finally, we analyze the cause of errors in estimation and investigate the effects of various parameters on the accuracy of estimation.
Join is an operation that is frequently used and the most expensive in processing database querie... more Join is an operation that is frequently used and the most expensive in processing database queries. In this paper we propose a new efficient join algorithm (called the MLGF-Join) for relations indexed by the tnul~ ilevrl grirlfile (MLGF), a multidimensional dynamic ...
We present a novel approach for computing link-based similarities among objects accurately by uti... more We present a novel approach for computing link-based similarities among objects accurately by utilizing the link information pertaining to the objects involved. We discuss the problems with previous link-based similarity measures and propose a novel approach for computing link based similarities that does not suffer from these problems. In the proposed approach each target object is represented by a vector. Each element of the vector corresponds to all the objects in the given data, and the value of each element denotes the weight for the corresponding object. As for this weight value, we propose to utilize the probability of reaching from the target object to the specific object, computed using the "Random Walk with Restart" strategy. Then, we define the similarity between two objects as the cosine similarity of the two vectors. In this paper, we provide examples to show that our approach does not suffer from the aforementioned problems. We also evaluate the performance o...
Because using social media has become a major part of people's daily lives, many of their per... more Because using social media has become a major part of people's daily lives, many of their personal characteristics are often implicitly or explicitly reflected in the content they share. We present a study of two personal characteristics-age and gender-related to user engagement on Instagram that can be determined through the characterization of images and tags. We demonstrate the strong influence of age and gender on Instagram use in terms of topical and content differences. We then build age and gender classification models that yield F1 scores of up to 88% and 74% in the detection of age and gender, respectively, and that better characterize users by images than by tags. We further demonstrate the robustness of our models using a new set of test data, with which the models exhibit greater overall performance than human raters. Our study highlights that future research should look to exploit images to a greater degree because they complement text and there are many unexamined ...
Proceedings of the 23rd International Conference on World Wide Web, 2014
Existing recommendation methods suffer from the data sparsity problem which means that most of us... more Existing recommendation methods suffer from the data sparsity problem which means that most of users have rated only a very small number of items, often resulting in low recommendation accuracy. In addition, for cold start users evaluating only few items, rating predictions with the methods also produce low accuracy. To address these problems, we propose a novel data imputation method that effectively substitutes missing ratings with probable values (i.e., imputed values). Our method successfully improves accuracy of recommendation methods from the following three aspects: (1) exploiting a trust network, (2) imputing only a part of missing ratings, and (3) applying them to any recommendation methods. Our method employs a bidirectional connection structure within a distance level for finding reliable users in exploiting a trust network as useful information. In addition, our method imputes only some missing ratings, called fillable ratings, whose imputed values are expected to be accurate with a sufficient level of confidence. Moreover, our imputation method is independent of, thus applicable to, any recommendation methods that may include application-specific ones and the most accurate one in each domain. We conduct experiments on three real-life datasets which arise from Epinions and Ciao. Our experimental results demonstrate that our method has recommendation accuracy better than existing recommendation methods equipped with imputation methods or trust networks, especially for cold start users.
Proceedings of the 2001 ACM symposium on Applied computing - SAC '01, 2001
This paper investigates the subsequence searching problem under time warping in sequence database... more This paper investigates the subsequence searching problem under time warping in sequence databases. Time warping enables to find sequences with similar changing patterns even when they are of different lengths. Our work is motivated by the observation that subsequence searches slow down quadratically as the total length of data sequences increases. To resolve this problem, we propose the Segment-Based Approach for Subsequence Searching Technique (SBASS), which modifies the similarity measure from time warping to piece-wise time warping and limits the number → → of possible subsequences to be compared with a query sequence. That is, the SBASS divides a data sequence X and a query sequence q into piece-wise → → → segments and compares q with only those subsequences which consist of n consecutive segments of X. Here, n is the number of segments in q. For efficient retrieval of similar subsequences, we extract feature vectors from all data segments exploiting their monotonically changing properties, and build a multi-dimensional index. Using this index, queries are processed with four steps: (1) index filtering, (2) feature filtering, (3) successor filtering, and (4) post-processing. The effectiveness of our approach is verified through experiments on synthetic data sets.
Proceedings of the ninth international conference on Information and knowledge management - CIKM '00, 2000
In this paper, w epropose a subsequence matching algorithm that supports normalization transform ... more In this paper, w epropose a subsequence matching algorithm that supports normalization transform in timeseries databases. Normalization transform enables nding sequences with similar uctuation patterns although they are not close to each other before the normalization transform. Application of the existing whole matching algorithm supporting normalization transform to the subsequence matching is feasible, but requires an index for ev ery possible length of the query sequence causing serious overhead on both storage space and update time. The proposed algorithm generates indexes only for a small number of di erent lengths of query sequences. F or subsequence matching it selects the most appropriate index among them. We can obtain better searc h performance by using more indexes. We c a l l o u r approach index interp olation. We formally pro ve t h a t the proposed algorithm does not cause false dismissal. F or performance evaluation, we h a ve conducted experiments using the indexes for only ve di erent lengths out of the lengths 256 512 of the query sequence. The results show that the proposed algorithm outperforms the sequential scan by up to 14.6 times on the average when the selectivity of the query is 10 ;5 .
Proceedings of the 20th ACM international conference on Information and knowledge management - CIKM '11, 2011
Color descriptors are one of the important features used in content-based image retrieval. The Do... more Color descriptors are one of the important features used in content-based image retrieval. The Dominant Color Descriptor (DCD) represents a few perceptually dominant colors in an image through color quantization. For image retrieval based on DCD, the earth mover's distance and the optimal color composition distance are proposed to measure the dissimilarity between two images. Although providing good retrieval results, both methods are too time-consuming to be used in a large image database. To solve the problem, we propose a new distance function that calculates an approximate earth mover's distance in linear time. To calculate the dissimilarity in linear time, the proposed approach employs the space-filling curve for multidimensional color space. To improve the accuracy, the proposed approach uses multiple curves and adjusts the color positions. As a result, our approach achieves order-of-magnitude time improvement but incurs small errors. We have performed extensive experiments to show the effectiveness and efficiency of the proposed approach. The results reveal that our approach achieves almost the same results with the EMD in linear time.
Proceedings of the 19th international conference on World wide web - WWW '10, 2010
In this paper, we propose a new paper ranking algorithm that gives a high rank to papers which is... more In this paper, we propose a new paper ranking algorithm that gives a high rank to papers which is credited by other authoritative papers or published in premier conferences or journals. Also, the proposed algorithm solves a problem that recent papers are rated poorly due to few citations.
The B +-tree was proposed to support sequential processing for the B-tree. To the extent of autho... more The B +-tree was proposed to support sequential processing for the B-tree. To the extent of authors' knowledge, however, there have been no studies supporting sequential processing in multidimensional point access methods(PAMs). To do this, the cells in a multilevel and multidimensional space managed by a multidimensional PAM must be linearly ordered systematically. In this paper, we present ideas for an approach that linearly orders cells in a multilevel and multidimensional space. Then, we present ideas for a novel sequential processing algorithm for region queries using this approach.
Proceedings of the 2011 ACM Symposium on Applied Computing - SAC '11, 2011
The blog world is a representative online society. To understand the nature of the blog world, th... more The blog world is a representative online society. To understand the nature of the blog world, there have been many research efforts on analyzing information diffusion and blogger activities. The independent cascade model is appropriate to explain information diffusion in the blog world. For the model to be employed, the blog world should be represented as a form of a network. For accurate analysis, it is crucial to assign a diffusion probability to each edge between a pair of bloggers in the blog network. In this paper, we propose a novel method to assign a diffusion probability to an edge for a pair of bloggers that reflects well the phenomenon of actual information diffusion between them. We verify the superiority of our approach by performing extensive experiments with real-world blog data.
In molecular biology, DNA sequence matching is one of the most crucial operations. Since DNA data... more In molecular biology, DNA sequence matching is one of the most crucial operations. Since DNA databases contain a huge volume of sequences, fast indexes are essential for efficient processing of DNA sequence matching. In this paper, we first point out the problems of the suffix tree, an index structure widely-used for DNA sequence matching, in the respects of the storage overhead, search performance, and difficulty in seamless integration with DBMS. Then, we propose a new index structure that resolves such problems. The proposed index structure consists of the two parts: the primary part realizes the trie as binary bit-string representation without any pointers, and the secondary part helps fast accesses of leaf nodes of the trie that need to be accessed for post-processing. We also suggest efficient algorithms based on that index for DNA sequence matching. To verify the superiority of the proposed approach, we conduct performance evaluation via a series of experiments. The results reveal that the proposed approach, which requires smaller storage space, can be a few orders of magnitude faster than the suffix tree.
This paper addresses the problem of timestamped event sequence matching, a new type of sequence m... more This paper addresses the problem of timestamped event sequence matching, a new type of sequence matching that retrieves the occurrences of interesting patterns from a timestamped event sequence. Timestamped event sequence matching is useful for discovering temporal causal relationships among timestamped events. In this paper, we first point out the shortcomings of prior approaches to this problem and then propose a novel method that employs an R *-tree to overcome them. To build an R *-tree, it places a time window at every position of a timestamped event sequence and represents each window as an n-dimensional rectangle by considering the first and last occurrence times of each event type. Here, n is the total number of disparate event types that may occur in a target application. When n is large, we apply a grouping technique to reduce the dimensionality of an R *-tree. To retrieve the occurrences of a query pattern from a timestamped event sequence, the proposed method first identifies a small number of candidates by searching an R *tree and then picks out true answers from them. We prove its robustness formally, and also show its effectiveness via extensive experiments.
2008 32nd Annual IEEE International Computer Software and Applications Conference, 2008
This paper addresses an approach that recommends investment types to stock investors by discoveri... more This paper addresses an approach that recommends investment types to stock investors by discovering useful rules from past changing patterns of stock prices in databases. First, we define a new rule model for recommending stock investment types. For a frequent pattern of stock prices, if its subsequent stock prices are matched to a condition of an investor, the model recommends a corresponding investment type for this stock. The frequent pattern is regarded as a rule head, and the subsequent part a rule body. We observed that the conditions on rule bodies are quite different depending on dispositions of investors while rule heads are independent of characteristics of investors in most cases. With this observation, we propose a new method that discovers and stores only the rule heads rather than the whole rules in a rule discovery process. This allows investors to impose various conditions on rule bodies flexibly, and also improves the performance of a rule discovery process by reducing the number of rules to be discovered. For efficient discovery and matching of rules, we propose methods for discovering frequent patterns, constructing a frequent pattern base, and its indexing. We also suggest a method that finds the rules matched to a query from a frequent pattern base, and a method that recommends an investment type by using the rules. Finally, we verify the effectiveness and the efficiency of our approach through extensive experiments with real-life stock data.
This paper designs and implements a taxi telematics service system, aiming at providing an effici... more This paper designs and implements a taxi telematics service system, aiming at providing an efficient framework by means of a Linux cluster to host emerging telematics services that need intensive computing. Combined with global positioning system and radio communication technology, the taxi telematics service system traces the position of taxis, finds a time saving route between start and destination points, dispatches the nearest taxi to the service call point based on the latest traffic information, and finally decides an efficient route for multiple destinations. The performance measurement result demonstrates that the implemented system can process up to 200 map matches for every minute, keeping average response time for other requests below 1.5 seconds.
The shape-based retrieval is defined as the operation that searches for the (sub)sequences whose ... more The shape-based retrieval is defined as the operation that searches for the (sub)sequences whose shapes are similar to that of a query sequence regardless of their actual element values. In this paper, we propose a similarity model suitable for shape-based retrieval and present an indexing method for supporting the similarity model. The proposed similarity model enables to retrieve similar shapes accurately by providing the combination of multiple shape-preserving transformations such as normalization, moving average, and time warping. Our indexing method stores every distinct subsequence concisely into the disk-based suffix tree for efficient and adaptive query processing. We allow the user to dynamically choose a similarity model suitable for a given application. More specifically, we allow the user to determine the parameter p of the distance function L p when submitting a query. The result of extensive experiments revealed that our approach not only successfully finds the subsequences whose shapes are similar to a query shape but also significantly outperforms the sequential scan method.
This paper discusses the way of processing time-series subsequence matching under time warping. T... more This paper discusses the way of processing time-series subsequence matching under time warping. Time warping enables sequences to be found with similar patterns even when they are of different lengths. The prefix-querying method is the first index-based approach that efficiently performs time-series subsequence matching under time warping without false dismissals. This method employs the L distance metric as a base distance function so as to allow users to issue queries conveniently. In this paper, we extend the prefix-querying method for absorbing L1, which is the most widely used as a base distance function in time-series subsequence matching under time warping, instead of L. We formally prove that the prefix-querying method with the L1 distance metric does not incur any false dismissals in the subsequence matching. To show its superiority, we conduct performance evaluation via a variety of experiments. The results reveal that our method achieves significant performance improvemen...
This paper discusses the effective processing of similarity search that supports time warping in ... more This paper discusses the effective processing of similarity search that supports time warping in large sequence databases. Time warping enables sequences with similar patterns to be found even when they are of different lengths. Prior methods for processing similarity search that supports time warping failed to employ multi-dimensional indexes without false dismissal since the time warping distance does not satisfy the triangular inequality. They have to scan the entire database, thus suffering from serious performance degradation in large databases. Another method that hires the suffix tree, which does not assume any distance function, also shows poor performance due to the large tree size. In this paper, we propose a novel method for similarity search that supports time warping. Our primary goal is to enhance the search performance in large databases without permitting any false dismissal. To attain this goal, we have devised a new distance function, D twÀlb ; which consistently underestimates the time warping distance and satisfies the triangular inequality. D twÀlb uses a 4-tuple feature vector that is extracted from each sequence and is invariant to time warping. For the efficient processing of similarity search, we employ a multi-dimensional index that uses the 4-tuple feature vector as indexing attributes, and D twÀlb as a distance function. We prove that our method does not incur false dismissal. To verify the superiority of our method, we have performed extensive experiments. The results reveal that our method achieves a significant improvement in speed up to 43 times faster with a data set containing real-world S&P 500 stock data sequences, and up to 720 times with data sets containing a very large volume of synthetic data sequences. The performance gain increases: (1) as the number of data sequences increases, (2) the average length of data sequences increases, and (3) as the tolerance in a query decreases. Considering the characteristics of real databases, these tendencies imply that our approach is suitable for practical applications.
As the total amount of traffic data in networks has been growing at an alarming rate, there is cu... more As the total amount of traffic data in networks has been growing at an alarming rate, there is currently a substantial body of research that attempts to mine traffic data with the purpose of obtaining useful information. For instance, there are some investigations into the detection of Internet worms and intrusions by discovering abnormal traffic patterns. However, since network traffic data contain information about the Internet usage patterns of users, network users' privacy may be compromised during the mining process. In this paper, we propose an efficient and practical method that preserves privacy during sequential pattern mining on network traffic data. In order to discover frequent sequential patterns without violating privacy, our method uses the N-repository server model, which operates as a single mining server and the retention replacement technique, which changes the answer to a query probabilistically. In addition, our method accelerates the overall mining process by maintaining the meta tables in each site so as to determine quickly whether candidate patterns have ever occurred in the site or not. Extensive experiments with real-world network traffic data revealed the correctness and the efficiency of the proposed method.
We propose a new dynamic method for multidimensional selectivity estimation for range queries tha... more We propose a new dynamic method for multidimensional selectivity estimation for range queries that works accurately independent of data distribution. Good estimation of selectivity is important for query optimization and physical database design. Our method employs the multilevel grid file (MLGF) for accurate estimation of multidimensional data distribution. The MLGF is a dynamic, hierarchical, balanced, multidimensional file structure that gracefully adapts to nonuniform and correlated distributions. We show that the MLGF directory naturally represents a multidimensional data distribution. We then extend it for further refinement and present the selectivity estimation method based on the MLGE Extensive experiments have been performed to test the accuracy of selectivity estimation. The results show that estimation errors are very small independent of distributions, even with correlated and/or highly skewed ones. Finally, we analyze the cause of errors in estimation and investigate the effects of various parameters on the accuracy of estimation.
Join is an operation that is frequently used and the most expensive in processing database querie... more Join is an operation that is frequently used and the most expensive in processing database queries. In this paper we propose a new efficient join algorithm (called the MLGF-Join) for relations indexed by the tnul~ ilevrl grirlfile (MLGF), a multidimensional dynamic ...
We present a novel approach for computing link-based similarities among objects accurately by uti... more We present a novel approach for computing link-based similarities among objects accurately by utilizing the link information pertaining to the objects involved. We discuss the problems with previous link-based similarity measures and propose a novel approach for computing link based similarities that does not suffer from these problems. In the proposed approach each target object is represented by a vector. Each element of the vector corresponds to all the objects in the given data, and the value of each element denotes the weight for the corresponding object. As for this weight value, we propose to utilize the probability of reaching from the target object to the specific object, computed using the "Random Walk with Restart" strategy. Then, we define the similarity between two objects as the cosine similarity of the two vectors. In this paper, we provide examples to show that our approach does not suffer from the aforementioned problems. We also evaluate the performance o...
Because using social media has become a major part of people's daily lives, many of their per... more Because using social media has become a major part of people's daily lives, many of their personal characteristics are often implicitly or explicitly reflected in the content they share. We present a study of two personal characteristics-age and gender-related to user engagement on Instagram that can be determined through the characterization of images and tags. We demonstrate the strong influence of age and gender on Instagram use in terms of topical and content differences. We then build age and gender classification models that yield F1 scores of up to 88% and 74% in the detection of age and gender, respectively, and that better characterize users by images than by tags. We further demonstrate the robustness of our models using a new set of test data, with which the models exhibit greater overall performance than human raters. Our study highlights that future research should look to exploit images to a greater degree because they complement text and there are many unexamined ...
Proceedings of the 23rd International Conference on World Wide Web, 2014
Existing recommendation methods suffer from the data sparsity problem which means that most of us... more Existing recommendation methods suffer from the data sparsity problem which means that most of users have rated only a very small number of items, often resulting in low recommendation accuracy. In addition, for cold start users evaluating only few items, rating predictions with the methods also produce low accuracy. To address these problems, we propose a novel data imputation method that effectively substitutes missing ratings with probable values (i.e., imputed values). Our method successfully improves accuracy of recommendation methods from the following three aspects: (1) exploiting a trust network, (2) imputing only a part of missing ratings, and (3) applying them to any recommendation methods. Our method employs a bidirectional connection structure within a distance level for finding reliable users in exploiting a trust network as useful information. In addition, our method imputes only some missing ratings, called fillable ratings, whose imputed values are expected to be accurate with a sufficient level of confidence. Moreover, our imputation method is independent of, thus applicable to, any recommendation methods that may include application-specific ones and the most accurate one in each domain. We conduct experiments on three real-life datasets which arise from Epinions and Ciao. Our experimental results demonstrate that our method has recommendation accuracy better than existing recommendation methods equipped with imputation methods or trust networks, especially for cold start users.
Proceedings of the 2001 ACM symposium on Applied computing - SAC '01, 2001
This paper investigates the subsequence searching problem under time warping in sequence database... more This paper investigates the subsequence searching problem under time warping in sequence databases. Time warping enables to find sequences with similar changing patterns even when they are of different lengths. Our work is motivated by the observation that subsequence searches slow down quadratically as the total length of data sequences increases. To resolve this problem, we propose the Segment-Based Approach for Subsequence Searching Technique (SBASS), which modifies the similarity measure from time warping to piece-wise time warping and limits the number → → of possible subsequences to be compared with a query sequence. That is, the SBASS divides a data sequence X and a query sequence q into piece-wise → → → segments and compares q with only those subsequences which consist of n consecutive segments of X. Here, n is the number of segments in q. For efficient retrieval of similar subsequences, we extract feature vectors from all data segments exploiting their monotonically changing properties, and build a multi-dimensional index. Using this index, queries are processed with four steps: (1) index filtering, (2) feature filtering, (3) successor filtering, and (4) post-processing. The effectiveness of our approach is verified through experiments on synthetic data sets.
Proceedings of the ninth international conference on Information and knowledge management - CIKM '00, 2000
In this paper, w epropose a subsequence matching algorithm that supports normalization transform ... more In this paper, w epropose a subsequence matching algorithm that supports normalization transform in timeseries databases. Normalization transform enables nding sequences with similar uctuation patterns although they are not close to each other before the normalization transform. Application of the existing whole matching algorithm supporting normalization transform to the subsequence matching is feasible, but requires an index for ev ery possible length of the query sequence causing serious overhead on both storage space and update time. The proposed algorithm generates indexes only for a small number of di erent lengths of query sequences. F or subsequence matching it selects the most appropriate index among them. We can obtain better searc h performance by using more indexes. We c a l l o u r approach index interp olation. We formally pro ve t h a t the proposed algorithm does not cause false dismissal. F or performance evaluation, we h a ve conducted experiments using the indexes for only ve di erent lengths out of the lengths 256 512 of the query sequence. The results show that the proposed algorithm outperforms the sequential scan by up to 14.6 times on the average when the selectivity of the query is 10 ;5 .
Proceedings of the 20th ACM international conference on Information and knowledge management - CIKM '11, 2011
Color descriptors are one of the important features used in content-based image retrieval. The Do... more Color descriptors are one of the important features used in content-based image retrieval. The Dominant Color Descriptor (DCD) represents a few perceptually dominant colors in an image through color quantization. For image retrieval based on DCD, the earth mover's distance and the optimal color composition distance are proposed to measure the dissimilarity between two images. Although providing good retrieval results, both methods are too time-consuming to be used in a large image database. To solve the problem, we propose a new distance function that calculates an approximate earth mover's distance in linear time. To calculate the dissimilarity in linear time, the proposed approach employs the space-filling curve for multidimensional color space. To improve the accuracy, the proposed approach uses multiple curves and adjusts the color positions. As a result, our approach achieves order-of-magnitude time improvement but incurs small errors. We have performed extensive experiments to show the effectiveness and efficiency of the proposed approach. The results reveal that our approach achieves almost the same results with the EMD in linear time.
Proceedings of the 19th international conference on World wide web - WWW '10, 2010
In this paper, we propose a new paper ranking algorithm that gives a high rank to papers which is... more In this paper, we propose a new paper ranking algorithm that gives a high rank to papers which is credited by other authoritative papers or published in premier conferences or journals. Also, the proposed algorithm solves a problem that recent papers are rated poorly due to few citations.
The B +-tree was proposed to support sequential processing for the B-tree. To the extent of autho... more The B +-tree was proposed to support sequential processing for the B-tree. To the extent of authors' knowledge, however, there have been no studies supporting sequential processing in multidimensional point access methods(PAMs). To do this, the cells in a multilevel and multidimensional space managed by a multidimensional PAM must be linearly ordered systematically. In this paper, we present ideas for an approach that linearly orders cells in a multilevel and multidimensional space. Then, we present ideas for a novel sequential processing algorithm for region queries using this approach.
Proceedings of the 2011 ACM Symposium on Applied Computing - SAC '11, 2011
The blog world is a representative online society. To understand the nature of the blog world, th... more The blog world is a representative online society. To understand the nature of the blog world, there have been many research efforts on analyzing information diffusion and blogger activities. The independent cascade model is appropriate to explain information diffusion in the blog world. For the model to be employed, the blog world should be represented as a form of a network. For accurate analysis, it is crucial to assign a diffusion probability to each edge between a pair of bloggers in the blog network. In this paper, we propose a novel method to assign a diffusion probability to an edge for a pair of bloggers that reflects well the phenomenon of actual information diffusion between them. We verify the superiority of our approach by performing extensive experiments with real-world blog data.
In molecular biology, DNA sequence matching is one of the most crucial operations. Since DNA data... more In molecular biology, DNA sequence matching is one of the most crucial operations. Since DNA databases contain a huge volume of sequences, fast indexes are essential for efficient processing of DNA sequence matching. In this paper, we first point out the problems of the suffix tree, an index structure widely-used for DNA sequence matching, in the respects of the storage overhead, search performance, and difficulty in seamless integration with DBMS. Then, we propose a new index structure that resolves such problems. The proposed index structure consists of the two parts: the primary part realizes the trie as binary bit-string representation without any pointers, and the secondary part helps fast accesses of leaf nodes of the trie that need to be accessed for post-processing. We also suggest efficient algorithms based on that index for DNA sequence matching. To verify the superiority of the proposed approach, we conduct performance evaluation via a series of experiments. The results reveal that the proposed approach, which requires smaller storage space, can be a few orders of magnitude faster than the suffix tree.
This paper addresses the problem of timestamped event sequence matching, a new type of sequence m... more This paper addresses the problem of timestamped event sequence matching, a new type of sequence matching that retrieves the occurrences of interesting patterns from a timestamped event sequence. Timestamped event sequence matching is useful for discovering temporal causal relationships among timestamped events. In this paper, we first point out the shortcomings of prior approaches to this problem and then propose a novel method that employs an R *-tree to overcome them. To build an R *-tree, it places a time window at every position of a timestamped event sequence and represents each window as an n-dimensional rectangle by considering the first and last occurrence times of each event type. Here, n is the total number of disparate event types that may occur in a target application. When n is large, we apply a grouping technique to reduce the dimensionality of an R *-tree. To retrieve the occurrences of a query pattern from a timestamped event sequence, the proposed method first identifies a small number of candidates by searching an R *tree and then picks out true answers from them. We prove its robustness formally, and also show its effectiveness via extensive experiments.
2008 32nd Annual IEEE International Computer Software and Applications Conference, 2008
This paper addresses an approach that recommends investment types to stock investors by discoveri... more This paper addresses an approach that recommends investment types to stock investors by discovering useful rules from past changing patterns of stock prices in databases. First, we define a new rule model for recommending stock investment types. For a frequent pattern of stock prices, if its subsequent stock prices are matched to a condition of an investor, the model recommends a corresponding investment type for this stock. The frequent pattern is regarded as a rule head, and the subsequent part a rule body. We observed that the conditions on rule bodies are quite different depending on dispositions of investors while rule heads are independent of characteristics of investors in most cases. With this observation, we propose a new method that discovers and stores only the rule heads rather than the whole rules in a rule discovery process. This allows investors to impose various conditions on rule bodies flexibly, and also improves the performance of a rule discovery process by reducing the number of rules to be discovered. For efficient discovery and matching of rules, we propose methods for discovering frequent patterns, constructing a frequent pattern base, and its indexing. We also suggest a method that finds the rules matched to a query from a frequent pattern base, and a method that recommends an investment type by using the rules. Finally, we verify the effectiveness and the efficiency of our approach through extensive experiments with real-life stock data.
This paper designs and implements a taxi telematics service system, aiming at providing an effici... more This paper designs and implements a taxi telematics service system, aiming at providing an efficient framework by means of a Linux cluster to host emerging telematics services that need intensive computing. Combined with global positioning system and radio communication technology, the taxi telematics service system traces the position of taxis, finds a time saving route between start and destination points, dispatches the nearest taxi to the service call point based on the latest traffic information, and finally decides an efficient route for multiple destinations. The performance measurement result demonstrates that the implemented system can process up to 200 map matches for every minute, keeping average response time for other requests below 1.5 seconds.
The shape-based retrieval is defined as the operation that searches for the (sub)sequences whose ... more The shape-based retrieval is defined as the operation that searches for the (sub)sequences whose shapes are similar to that of a query sequence regardless of their actual element values. In this paper, we propose a similarity model suitable for shape-based retrieval and present an indexing method for supporting the similarity model. The proposed similarity model enables to retrieve similar shapes accurately by providing the combination of multiple shape-preserving transformations such as normalization, moving average, and time warping. Our indexing method stores every distinct subsequence concisely into the disk-based suffix tree for efficient and adaptive query processing. We allow the user to dynamically choose a similarity model suitable for a given application. More specifically, we allow the user to determine the parameter p of the distance function L p when submitting a query. The result of extensive experiments revealed that our approach not only successfully finds the subsequences whose shapes are similar to a query shape but also significantly outperforms the sequential scan method.
This paper discusses the way of processing time-series subsequence matching under time warping. T... more This paper discusses the way of processing time-series subsequence matching under time warping. Time warping enables sequences to be found with similar patterns even when they are of different lengths. The prefix-querying method is the first index-based approach that efficiently performs time-series subsequence matching under time warping without false dismissals. This method employs the L distance metric as a base distance function so as to allow users to issue queries conveniently. In this paper, we extend the prefix-querying method for absorbing L1, which is the most widely used as a base distance function in time-series subsequence matching under time warping, instead of L. We formally prove that the prefix-querying method with the L1 distance metric does not incur any false dismissals in the subsequence matching. To show its superiority, we conduct performance evaluation via a variety of experiments. The results reveal that our method achieves significant performance improvemen...
This paper discusses the effective processing of similarity search that supports time warping in ... more This paper discusses the effective processing of similarity search that supports time warping in large sequence databases. Time warping enables sequences with similar patterns to be found even when they are of different lengths. Prior methods for processing similarity search that supports time warping failed to employ multi-dimensional indexes without false dismissal since the time warping distance does not satisfy the triangular inequality. They have to scan the entire database, thus suffering from serious performance degradation in large databases. Another method that hires the suffix tree, which does not assume any distance function, also shows poor performance due to the large tree size. In this paper, we propose a novel method for similarity search that supports time warping. Our primary goal is to enhance the search performance in large databases without permitting any false dismissal. To attain this goal, we have devised a new distance function, D twÀlb ; which consistently underestimates the time warping distance and satisfies the triangular inequality. D twÀlb uses a 4-tuple feature vector that is extracted from each sequence and is invariant to time warping. For the efficient processing of similarity search, we employ a multi-dimensional index that uses the 4-tuple feature vector as indexing attributes, and D twÀlb as a distance function. We prove that our method does not incur false dismissal. To verify the superiority of our method, we have performed extensive experiments. The results reveal that our method achieves a significant improvement in speed up to 43 times faster with a data set containing real-world S&P 500 stock data sequences, and up to 720 times with data sets containing a very large volume of synthetic data sequences. The performance gain increases: (1) as the number of data sequences increases, (2) the average length of data sequences increases, and (3) as the tolerance in a query decreases. Considering the characteristics of real databases, these tendencies imply that our approach is suitable for practical applications.
As the total amount of traffic data in networks has been growing at an alarming rate, there is cu... more As the total amount of traffic data in networks has been growing at an alarming rate, there is currently a substantial body of research that attempts to mine traffic data with the purpose of obtaining useful information. For instance, there are some investigations into the detection of Internet worms and intrusions by discovering abnormal traffic patterns. However, since network traffic data contain information about the Internet usage patterns of users, network users' privacy may be compromised during the mining process. In this paper, we propose an efficient and practical method that preserves privacy during sequential pattern mining on network traffic data. In order to discover frequent sequential patterns without violating privacy, our method uses the N-repository server model, which operates as a single mining server and the retention replacement technique, which changes the answer to a query probabilistically. In addition, our method accelerates the overall mining process by maintaining the meta tables in each site so as to determine quickly whether candidate patterns have ever occurred in the site or not. Extensive experiments with real-world network traffic data revealed the correctness and the efficiency of the proposed method.
Uploads
Papers by Sang-wook Kim