• Annual Conference on Human and Language Technology
• /
• 2010.10a
• /
• pp.68-72
• /
• 2010

본 논문에서는 그래프 기반의 최대신장트리(Maximum Spanning Tree)를 이용한 한국어 의존구문분석 방법을 제안한다. 우리는 최대신장트리 알고리즘을 한국어의 특성인 지배성분 후위의 원칙과 투사성의 원칙을 적용하여 한국어 의존구문분석에 적합한 알고리즘을 만들었다. 제안한 알고리즘은 기존의 한국어 의존구문분석의 방법들보다 낮은 시간복잡도를 가지며 대용량 말뭉치를 학습하기 위해 증분학습이 가능하고 비교적 학습속도가 빠른 Averaged Perceptron 알고리즘을 사용하였다. 실험결과 제안한 방법은 비교적 열악한 환경인 복문이 포함된 장문의 문장에서도 뛰어난 성능을 보여주었다,

• Journal of the Korea Society of Computer and Information
• /
• v.20 no.3
• /
• pp.107-113
• /
• 2015

This paper presents a polynomial time algorithm to the minimum cardinality feedback edge set and minimum weight feedback edge set problems. The algorithm makes use of the property wherein the sum of the minimum spanning tree edge set and the minimum feedback edge set equals a given graph's edge set. In other words, the minimum feedback edge set is inherently a complementary set of the former. The proposed algorithm, in pursuit of the optimal solution, modifies the minimum spanning tree finding Kruskal's algorithm so as to arrange the weight of edges in a descending order and to assign cycle-deficient edges to the maximum spanning tree edge set MXST and cycle-containing edges to the feedback edge set FES. This algorithm runs with linear time complexity, whose execution time corresponds to the number of edges of the graph. When extensively tested on various undirected graphs both with and without the weighed edge, the proposed algorithm has obtained the optimal solutions with 100% success and accuracy.

• Journal of KIISE:Computer Systems and Theory
• /
• v.30 no.5_6
• /
• pp.319-323
• /
• 2003

Let P be a set of n points in the plane. A minimum spanning tree(MST) is a spanning tree connecting n points of P such that the sum of lengths of edges of the tree is minimized. A diameter of a tree is the maximum length of paths connecting two points of a spanning tree of P. The problem considered in this paper is to compute the spanning tree whose diameter is minimized over all spanning trees of P. We call such tree a minimum-diameter spanning tree(MDST). The best known previous algorithm[3] finds MDST in $O(n^2)$ time. In this paper, we suggest an approximation algorithm to compute a spanning tree whose diameter is no more than 5/4 times that of MDST, running in O(n$^2$log$^2$n) time. This is the first approximation algorithm on the MDST problem.

• The KIPS Transactions:PartA
• /
• v.17A no.2
• /
• pp.103-112
• /
• 2010

Cluster sensor network is a sensor network where input nodes crowd densely around some nuclei. Steiner minimum tree is a tree connecting all input nodes with introducing some additional nodes called Steiner points. This paper proposes a mechanism for efficient construction of a cluster sensor network connecting all sensor nodes and base stations using connections between nodes in each belonged cluster and between every cluster, and using repetitive constructions of approximate Steiner minimum trees. In experiments, while taking 1170.5% percentages more time to build cluster sensor network than the method of Euclidian minimum spanning tree, the proposed mechanism whose time complexity is O($N^2$) could spend only 20.3 percentages more time for building 0.1% added length network in comparison with the method of Euclidian minimum spanning tree. The mechanism could curtail the built trees' average length by maximum 3.7 percentages and by average 1.9 percentages, compared with the average length of trees built by Euclidian minimum spanning tree method.

• The KIPS Transactions:PartA
• /
• v.17A no.5
• /
• pp.213-220
• /
• 2010

In this paper, a mechanism connecting all input edges with minimum length through Steiner tree is proposed. Edges are convertible into communication lines, roads, railroads or trace of moving object. Proposed mechanism could be applied to connect these edges with minimum cost. In our experiments where input edge number and maximum connections per edge are used as input parameters, our mechanism made connection length decrease average 6.8%, while building time for a connecting solution increase average 192.0% comparing with the method using minimum spanning tree. The result shows our mechanism might be well applied to the applications where connecting cost is more important than building time for a connecting solution.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.14 no.1
• /
• pp.51-59
• /
• 2014

This paper suggests a method of lessening number of a graph's edges population in order to rapidly obtain the minimum spanning tree. The present minimum spanning tree algorithm works on all the edges of the graph. However, the suggested algorithm reduces the edges population size by means of applying a method of deleting maximum weight edges in advance from vertices with more than 2 valencies. Next, it applies a stopping criterion which ideally terminates Borůvka, Prim, Kruskal and Reverse-Delete algorithms for reduced edges population. On applying the suggested algorithm to 9 graphs, it was able to minimize averagely 83% of the edges that do not become MST. In addition, comparing to the original graph, edges are turned out to be lessened 38% by Borůvka, 37% by Prim, 39% by Kruskal and 73% by Reverse-Delete algorithm, and thereby the minimum spanning tree is obtained promptly.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.14 no.2
• /
• pp.35-42
• /
• 2014

This paper suggests a method that obtains the minimum spanning tree (MST) far more easily and rapidly than the present ones. The suggested algorithm, firstly, simplifies a graph by means of reducing the number of edges of the graph. To achieve this, it applies a method of eliminating the maximum weight edge if the valency of vertices of the graph is equal to or more than 3. As a result of this step, we can obtain the reduced edge population. Next, it applies a method in which the maximum weight edge is eliminated within the cycle. On applying the suggested population minimizing and maximum weight edge deletion algorithms to 9 various graphs, as many as the number of cycles of the graph is executed and MST is easily obtained. It turns out to lessen 66% of the number of cycles and obtain the MST in at least 2 and at most 8 cycles by only deleting the maximum weight edges.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.13 no.2
• /
• pp.103-114
• /
• 2013

This paper suggests a method to reduce the number of performances of Kruskal and Reverse-delete algorithms. Present Kruskal and Reverse-delete algorithms verify whether the cycle occurs within the edges of the graph. For this reason, they have problems of unnecessarily performing extra algorithms from the edges, even though they've already obtained the minimum spanning tree. This paper, first of all, suggests the 1st method which reduces the no. of performances by introducing stop point criteria of algorithm, but at the same time, performs algorithms from all the edges, just like how Kruskal and Reverse-delete algorithms. Next, it suggests the 2nd method which finds the minimum spanning tree from the remaining edges after getting rid of all the unnecessary edges which are considered not to affect the minimum spanning tree. These suggested methods have an effect of terminating algorithm at least 1.4 times and at most 3.86times than Kruskal and Reverse-delete algorithms, when applied to the real graphs. We have found that the 2nd method of the Reverse-delete algorithm has the fastest speed in terminating an algorithm, among 4 algorithms which are results of the 2 suggested methods being applied to 2 algorithms.

• Journal of KIISE
• /
• v.44 no.11
• /
• pp.1236-1243
• /
• 2017

In this paper, we propose a novel way of producing keyword networks, named LSI-based ClusterTextRank, which extracts significant key words from a set of clusters with a mutual information metric, and constructs an association network using latent semantic indexing (LSI). The proposed method reduces the dimension of documents through LSI, decomposes documents into multiple clusters through k-means clustering, and expresses the words within each cluster as a maximal spanning tree graph. The significant key words are identified by evaluating their mutual information within clusters. Then, the method calculates the similarities between the extracted key words using the term-concept matrix, and the results are represented as a keyword association network. To evaluate the performance of the proposed method, we used travel-related blog data and showed that the proposed method outperforms the existing TextRank algorithm by about 14% in terms of accuracy.

• The KIPS Transactions:PartA
• /
• v.18A no.5
• /
• pp.215-224
• /
• 2011

In this paper, a mechanism connecting all weighted migration routes with minimum cost with EGOSST is proposed. Weighted migration routes may be converted to weighted input edges considered as not only traces but also traffics or trip frequencies of moving object on communication lines, roads or railroads. Proposed mechanism can be used in more wide and practical area than mechanisms considering only moving object traces. In our experiments, edge number, maximum weight for input edges, and detail level for grid are used as input parameters. The mechanism made connection cost decrease average 1.07% and 0.43% comparing with the method using weight minimum spanning tree and weight steiner minimum tree respectively. When grid detail level is 0.1 and 0.001, while each execution time for a connecting solution increases average 97.02% and 2843.87% comparing with the method using weight minimum spanning tree, connecting cost decreases 0.86% and 1.13% respectively. This shows that by adjusting grid detail level, proposed mechanism might be well applied to the applications where designer must grant priority to reducing connecting cost or shortening execution time as well as that it can provide good solutions of connecting migration routes with weights.