• 한국정보과학회논문지:컴퓨팅의 실제 및 레터
• /
• 제16권1호
• /
• pp.110-114
• /
• 2010

본 논문에서는 TextRank 알고리즘을 이용한 문서 범주화 방법에 대해 기술한다. TextRank 알고리즘은 그래프 기반의 순위화 알고리즘이다. 문서에서 나타나는 각각의 단어를 노드로, 단어들 사이의 동시출현성을 이용하여 간선을 만들면 문서로부터 그래프를 생성할 수 있다. TextRank 알고리즘을 이용하여 생성된 그래프로부터 중요도가 높은 단어를 선택하고, 그 단어와 인접한 단어를 묶어 하나의 자질로 사용하여 문서 분류를 수행하였다. 동시출현 자질(인접한 단어 쌍)은 단어 하나가 갖는 의미를 보다 명확하게 만들어주므로 문서 분류에 좋은 자질로 사용될 수 있을 것이라 가정하였다. 문서 분류기로는 지지 벡터 기계, 베이지언 분류기, 최대 엔트로피 모델, k-NN 분류기 등을 사용하였다. 20 Newsgroups 문서 집합을 사용한 실험에서 모든 분류기에서 제안된 방법을 사용했을 때, 문서 분류 성능이 향상된 결과를 확인할 수 있었다.

• 한국통신학회논문지
• /
• 제16권4호
• /
• pp.319-325
• /
• 1991

이 논문에서는 일반화된 관측 모델의 특수한 때에 순위 통계량을 써서 알려진 신호를 비모수 검파하는 한 가지 방법을 생각하였다. 좀더 구체적으로는 신호의존성 잡음 모델에서 알려진 신호 국소 최적 순위 검파기를 얻고 이를 순가산성 잡음 모델에서 얻은 국소 최적 순위 검파기와 견주어 보았다. 또한 국소 최적 순위 검파기의 검정 통계량을 이루는 점수 함수의 몇가지 보기를 보였다.

• ETRI Journal
• /
• 제38권1호
• /
• pp.52-61
• /
• 2016

A microblog is a service typically offered by online social networks, such as Twitter and Facebook. From the perspective of information dissemination, we define the concept behind a spreading matrix. A new WeiboRank algorithm for identification of key nodes in microblog networks is proposed, taking into account parameters such as a user's direct appeal, a user's influence region, and a user's global influence power. To investigate how measures for ranking influential users in a network correlate, we compare the relative influence ranks of the top 20 microblog users of a university network. The proposed algorithm is compared with other algorithms - PageRank, Betweeness Centrality, Closeness Centrality, Out-degree - using a new tweets propagation model - the Ignorants-Spreaders-Rejecters model. Comparison results show that key nodes obtained from the WeiboRank algorithm have a wider transmission range and better influence.

• Communications for Statistical Applications and Methods
• /
• 제16권1호
• /
• pp.115-125
• /
• 2009

경영.경제분야에서 사용되는 모형 가운데 연립방정식 모형은 모형 내에서 결정되는 내생변수와 모형 외부로부터 결정된 외생변수들로 구성된 M개의 방정식과 T개의 관찰치로 이루어진 회귀방정식체계이며, 모형에 대한 모수식별 및 유일해의 존재여부에 대한 결정방법으로 순서조건과 계수조건이 있다. 그러나 대부분 연립방정식 모형이 이들 조건을 만족한다는 가정하에서 모수들을 추정하기 때문에 추정값이 비효율적이거나, 유일한 모수 추정값이 존재하지 않는 경우가 이들 조건에 따라 발생할 수 있다. 본 연구에서는 순서조건을 만족한다는 가정 하에서 계수조건의 충족여부를 검정하기 위한 검정통계량을 새롭게 제시하고 이의 근사분포를 도출하였으며, 이와 함께 모의 실험을 통하여 제안한 검정통계량의 검정력을 살펴보았다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권1호
• /
• pp.109-134
• /
• 2018

Collaborative Filtering (CF) is widely used in recommendation field, which can be divided into rating-based CF and learning-to-rank based CF. Although many methods have been proposed based on these two kinds of CF, there still be room for improvement. Firstly, the data sparsity problem still remains a big challenge for CF algorithms. Secondly, the malicious rating given by some illegal users may affect the recommendation accuracy. Existing CF algorithms seldom took both of the two observations into consideration. In this paper, we propose a recommendation method based on listwise learning-to-rank by incorporating users' social information. By taking both ratings and order of items into consideration, the Plackett-Luce model is presented to find more accurate similar users. In order to alleviate the data sparsity problem, the improved matrix factorization model by integrating the influence of similar users is proposed to predict the rating. On the basis of exploring the trust relationship between users according to their social information, a listwise learning-to-rank algorithm is proposed to learn an optimal ranking model, which can output the recommendation list more consistent with the user preference. Comprehensive experiments conducted on two public real-world datasets show that our approach not only achieves high recommendation accuracy in relatively short runtime, but also is able to reduce the impact of malicious ratings.

• ETRI Journal
• /
• 제41권2호
• /
• pp.235-241
• /
• 2019

In this paper, we propose a rank-weighted reconstruction feature to improve the robustness of a feed-forward deep neural network (FFDNN)-based acoustic model. In the FFDNN-based acoustic model, an input feature is constructed by vectorizing a submatrix that is created by slicing the feature vectors of frames within a context window. In this type of feature construction, the appropriate context window size is important because it determines the amount of trivial or discriminative information, such as redundancy, or temporal context of the input features. However, we ascertained whether a single parameter is sufficiently able to control the quantity of information. Therefore, we investigated the input feature construction from the perspectives of rank and nullity, and proposed a rank-weighted reconstruction feature herein, that allows for the retention of speech information components and the reduction in trivial components. The proposed method was evaluated in the TIMIT phone recognition and Wall Street Journal (WSJ) domains. The proposed method reduced the phone error rate of the TIMIT domain from 18.4% to 18.0%, and the word error rate of the WSJ domain from 4.70% to 4.43%.

• 대한임베디드공학회논문지
• /
• 제10권2호
• /
• pp.65-71
• /
• 2015

It has been known that GMM (Gaussian Mixture Model) based speaker identification systems using ML (Maximum Likelihood) and WMR (Weighting Model Rank) demonstrate very high performances. However, such systems are not so effective under practical environments, in terms of real time processing, because of their high calculation costs. In this paper, we propose a new speaker-pruning algorithm that effectively reduces the calculation cost. In this algorithm, we select 20% of speaker models having higher likelihood with a part of input speech and apply MWMR (Modified Weighted Model Rank) to these selected speaker models to find out identified speaker. To verify the effectiveness of the proposed algorithm, we performed speaker identification experiments using TIMIT database. The proposed method shows more than 60% improvement of reduced processing time than the conventional GMM based system with no pruning, while maintaining the recognition accuracy.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권9호
• /
• pp.3745-3761
• /
• 2020

Low-rank representation methods already achieve many applications in the image reconstruction. However, for high-gradient image patches with rich texture details and strong edge information, it is difficult to find sufficient similar patches. Existing low-rank representation methods usually destroy image critical details and fail to preserve edge structure. In order to promote the performance, a new representation-based image super-resolution reconstruction method is proposed, which combines gradient domain guided image filter with the structure-constrained low-rank representation so as to enhance image details as well as reveal the intrinsic structure of an input image. Firstly, we extract the gradient domain guided filter of each atom in high resolution dictionary in order to acquire high-frequency prior information. Secondly, this prior information is taken as a structure constraint and introduced into the low-rank representation framework to develop a new model so as to maintain the edges of reconstructed image. Thirdly, the approximate optimal solution of the model is solved through alternating direction method of multipliers. After that, experiments are performed and results show that the proposed algorithm has higher performances than conventional state-of-the-art algorithms in both quantitative and qualitative aspects.

• Asia pacific journal of information systems
• /
• 제25권3호
• /
• pp.548-562
• /
• 2015

The growing popularity of mobile application has led to researchers and practitioners needing to understand users' mobile application download behaviors. Using large-scale transaction data obtained from a leading Korean telecommunications company, we empirically explore how application download rank, which appears to users when they decide to download a new application, affects their extended mobile application download. This terminology refers to downloading an additional application in the same category as those that they have already downloaded. We also consider IT characteristics, user characteristics, and application type that might be associated with the extended application download. The analysis generates the following result. Overall, a higher rank of a new application encourages the extended application download, but the linear relationship between the rank and the extended application download disappears when critical rank points are incorporated into the model. Further, no quadratic effect of rank is found in the extended application download. Based on the results, we suggest theoretical and managerial implications.

• Communications for Statistical Applications and Methods
• /
• 제9권1호
• /
• pp.101-113
• /
• 2002

In this paper we introduce a new score generating (unction for the rank regression in the linear regression model. The score function compares the $\gamma$'th and s\`th power of the tail probabilities of the underlying probability distribution. We show that the rank estimate asymptotically converges to a multivariate normal. further we derive the asymptotic Pitman relative efficiencies and the most efficient values of $\gamma$ and s under the symmetric distribution such as uniform, normal, cauchy and double exponential distributions and the asymmetric distribution such as exponential and lognormal distributions respectively.