Min-Hee Jang

Trust prediction using positive, implicit, and negative information

We propose a novel method to predict accurately trust relationships of a target user even if he/she does not have much interaction information. The proposed method considers positive, implicit, and negative information of all users in a network based on belief propagation to predict trust relationships of a target user.

Performance of TPR*-Trees for Predicting Future Positions of Moving Objects in U-Cities

The TPR*-tree is the most widely-used index structure for effectively predicting the future positions of moving objects. The TPR*-tree, however, has the problem that both of the dead spacein a bounding region and the overlap among bounding regions become larger as the prediction time point in the future gets farther. This makes more nodes within the TPR*-tree accessed in query processing time, which incurs serious performance degradation. In this paper, we examine the performance problem quantitatively via a series of experiments. First, we show how much the performance deteriorates as a prediction time point gets farther from the present, and also show how the frequent updates of positions of moving objects alleviate this problem. Our contribution would help provide important clues to devise strategies improving the performance of TPR*-trees further.

Shader space navigator: a turbo for an intuitive and effective shading process

In this paper, we first point out difficulties faced by CG artists in the shading process: (1) a lot of technical details on shaders required, (2) long rendering time, and (3) repeated cumbersome trial-and-errors. To make them overcome such difficulties, we propose Shader Space Navigator, a system that efficiently searches for shaders similar to a given query shader. With Shader Space Navigator, CG artists find quality shaders from the database that are very close to the final result shader, and thus complete the shading process easily by slightly tuning some attributes of those shaders. As a result, the CG artists can create their final shaders in an intuitive and efficient way thereby avoiding a large number of time-consuming rendering processes.

Approximate k-Nearest Neighbor Search Based on the Earth Mover's Distance for Efficient Content-based Information Retrieval

The Earth Movers Distance (EMD) is one of the most-widely used distance functions to measure the similarity between two multimedia objects. While providing good search results, the EMD is too much time-consuming to be used in large multimedia databases. To solve the problem, we propose an approximate k-nearest neighbor (k-NN) search method based on the EMD. First, the proposed method builds an index using the M-tree, a distance-based multi-dimensional index structure, to reduce the disk access overhead. When building the index, we reduce the number of features in the multimedia objects through dimensionality-reduction. When performing the k-NN search on the M-tree, we find a small set of candidates from the disk using the index and then perform the post-processing on them. Second, the proposed method uses the approximate EMD for index retrieval and post-processing to reduce the computational overhead of the EMD. To compensate the errors due to the approximation, the method provides a way of accuracy improvement of the approximate EMD. We performed extensive experiments to show the efficiency of the proposed method.

A semantic similarity measure in document databases: an earth mover's distance-based approach

Measuring document similarity is important in order to find documents which are similar to a given query document from a user. Text-based document similarity is measured by comparing the words in two documents. The representative text-based document similarity is the cosine similarity. Since the cosine similarity computes document similarity by estimating the frequency of common words, it cannot reflect word similarity. To solve this problem, we propose a new document similarity measure based on the earth movers distance (EMD). The EMD is one of the most popular distance functions used to search similar multimedia contents and is known to provide good search results. To apply the EMD to compute document similarity, we have to solve two problems: (1) The EMD is too time consuming to be used in a document database, (2) the distance between words should be defined. Our proposed approach first extracts topics as new features of a document by applying the latent Dirichlet allocation, which is a generative model of a document. It can decrease the computational cost of the EMD because the number of topics is much smaller than the number of words in a document. After extracting the topics, the proposed approach calculates the distance between topics based on the relation between the topics and the words in a document database, thereby making computing document similarity based on the EMD possible. Our approach searches documents more accurately since we can consider the semantic similarity by using the EMD. Experimental results on a real-world document database indicate that the proposed approach outperforms the cosine similarity in terms of the accuracy and the performance.

Hierarchical data organization for effective retrieval of similar shaders

Generating a shader is an important factor for portraying realistic objects on the screen using 3D computer graphics. However, since generating a shader requires significant amounts of time and effort, the user must be able to search for a desired shader from a large shader database. This paper proposes a novel hierarchical clustering algorithm that structuralizes a shader database by clustering similar shaders in order to facilitate the shader search process. Since conventional hierarchical clustering methods do not take into account the cluster diameter and the number of shaders in a cluster, the resulting databases structure is highly likely to be a skewed tree. The proposed method represents a shader database into a graph and recursively uses a graph partitioning algorithm to perform hierarchical clustering. In this process, since the cluster diameter and the number of shaders in a cluster are taken into account preventing a single cluster from becoming excessively large, we can construct a hierarchical database structure close to a balanced state. Various experiments were conducted to verify the merits of the proposed method. The experimental results indicate that the proposed method provides a high level of accuracy as well as a structure close to a balanced state.

Active Adjustment: An Approach for Improving the Search Performance of the TPR * -tree

Recently, with the advent of applications using locations of moving objects, it becomes crucial to develop efficient index schemes for spatio-temporal databases. The -tree is most popularly accepted as an index structure for processing future-time queries. In the -tree, the future locations of moving objects are predicted based on the CBR(Conservative Bounding Rectangle). Since the areas predicted from CBRs tend to grow rapidly over time, CBRs thus enlarged lead to serious performance degradation in query processing. Against the problem, we propose a new method to adjust CBRs to be tight, thereby improving the performance of query processing. Our method examines whether the adjustment of a CBR is necessary when accessing a leaf node for processing a user query. Thus, it does not incur extra disk I/Os in this examination. Also, in order to make a correct decision, we devise a cost model that considers both the I/O overhead for the CBR adjustment and the performance gain in the future-time owing to the CBR adjustment. With the cost model, we can prevent unusual expansions of BRs even when updates on nodes are infrequent and also avoid unnecessary execution of the CBR adjustment. For performance evaluation, we conducted a variety of experiments. The results show that our method improves the performance of the original -tree significantly.

Active adjustment: an approach for improving the performance of the TPR*-tree

The TPR*-tree is most popularly accepted as an index structure for processing future-time queries in moving object databases. In the TPR*-tree, the future locations of moving objects are predicted based on the CBR(Conservative Bounding Rectangle). Since the areas predicted from CBRs tend to grow rapidly over time, CBRs thus enlarged lead to serious performance degradation in query processing. Against the problem, we propose a novel method to adjust CBRs to be tight, thereby improving the performance of query processing. Our method examines whether the adjustment of a CBR is necessary when accessing a leaf node for processing a user query. Thus, it does not incur extra disk I/Os in this examination. Also, in order to make a correct decision, we devise a cost model that considers the I/O overhead for the CBR adjustment and the performance gain in the future-time owing to the CBR adjustment. With the cost model, we can prevent unusual expansions of BRs even when updates on nodes are infrequent and also avoid unnecessary execution of the CBR adjustment. For performance evaluation, we conducted a variety of experiments. The results show that our method improves the performance of the original TPR*-tree significantly.

Generating trajectories on road networks

Recently, researches are being in progress using the trajectories of moving objects. Most researches usually used data generated by trajectory generators since it is difficult to obtain a trajectory data set of moving objects in real world. Most previous trajectory generators created trajectories of objects moving over Euclidean space, and therefore they can not be directly applied to road network environment. In this paper, we propose a method for generating trajectories of objects moving over road networks. To generate trajectories, we consider the most important characteristic of network-based moving objects that in real world most objects move on given networks with the shortest path from a starting point to a destination. The trajectory data set of moving objects which is generated by the proposed method can be used in various applications such as location-based services since it reflects the users driving preference on real network environments.