Dojin Choi

An efficient continuous subgraph matching scheme over graph streams

In many applications, a graph is often utilized to represent an interrelation among objects. Especially, as IoT and social network services, which represent relationships among objects as a graph, continue to grow, many studies has been conducted to handle graph streams that are generated in real time. In this paper, we propose a continuous subgraph matching scheme based on the query decomposition considering the node degree in graph streams. In order to reuse the index, we decompose the query based on the degree information and then manage the decomposed subqueries as index. The decomposed subqueries are reused or decomposed again by the generated query. The proposed scheme also reuses the existing computed results through the decomposed indexes, which results in reducing the load of query processing. We demonstrate the superiority of the proposed scheme through various performance evaluations in real world datasets.

Academic Expert Search Method Using Importance and Quality of Papers

An expert search method using a large amount of academic data that can provide users with representative research results and advice is required. Since the existing expert search methods perform the expert search based on user profile or activity information, they have a problem that it is hard to discriminate the expert when we do not know the user profile or activity information. In this paper, we propose an academic expert search method using the importance and quality of a paper. The importance of a paper is computed by considering its scarcity and up-to-date topics. The quality of a paper is evaluated by considering the number of citations, IF of Journal, recency and author relations. To show the superiority of the proposed method, we compare it with the existing scheme through the performance evaluation in terms of recall and precision