• Annual Conference on Human and Language Technology
• /
• 2013.10a
• /
• pp.18-23
• /
• 2013

기존 대화시스템과 달리 대화형 개인 비서 시스템은 사용자에게 정보를 제공하기 위해 앱(APP)을 구동하는 방법을 사용한다. 사용자가 앱을 통해 정보를 얻고자 할 때, 사용자가 필요로 하는 정보를 제공해주기 위해서는 사용자의 목적을 정확하게 인식하는 작업이 필요하다. 그 작업 중 중요한 두 요소는 개체명 인식과 문장목적 인식이다. 문장목적 인식이란, 사용자의 문장을 분석해 하나의 앱에 존재하는 여러 정보 중 사용자가 원하는 정보(문장의 목적)가 무엇인지 찾아주는 인식작업이다. 이러한 인식시스템을 구축하는 방법 중 대표적인 방법은 사전규칙방법과 기계학습방법이다. 사전규칙은 사전정보와 규칙을 적용하는 방법으로, 시간이 지남에 따라 새로운 규칙을 추가해야하는 문제가 있으며, 규칙이 일반화되지 않을 경우 오류가 증가하는 문제가 있다. 또 두 인식작업을 파이프라인 방식으로 적용 할 경우, 개체명 인식단계에서의 오류를 가지고 문장목적 인식단계로 넘어가기 때문에 두 단계에 걸친 성능저하와 속도저하를 초래할 수 있다. 이러한 문제점을 해결하기 위해 우리는 통계기반의 기계학습방법인 Conditional Random Fields(CRF)를 사용한다. 또한 사전정보를 CRF와 결합함으로써, 단독으로 수행하는 CRF방식의 성능을 개선시킨다. 개체명과 문장목적인식의 구조를 분석한 결과, 비슷한 자질을 사용할 수 있다고 판단하여, 두 작업을 동시에 수행하는 방법을 제안한다. 실험결과, 사전규칙방법보다 제안한 방법이 문장단위 2.67% 성능개선을 보였다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.9
• /
• pp.4814-4832
• /
• 2019

Plagiarism detection is increasingly exploiting text alignment. Text alignment involves extracting the plagiarism passages in a pair of the suspicious document and its source document. The heuristics have achieved excellent performance in text alignment. However, the further improvements of the heuristic methods mainly depends more on the experiences of experts, which makes the heuristics lack of the abilities for continuous improvements. To address this problem, machine learning maybe a proper way. Considering the position relations and the context of text segments pairs, we formalize the text alignment task as a problem of sequence labeling, improving the current methods at the model level. Especially, this paper proposes to use the probabilistic graphical model to tag the observed sequence of pairs of text segments. Hence we present the sequence labeling approach for text alignment in plagiarism detection based on Conditional Random Fields. The proposed approach is evaluated on the PAN@CLEF 2012 artificial high obfuscation plagiarism corpus and the simulated paraphrase plagiarism corpus, and compared with the methods achieved the best performance in PAN@CLEF 2012, 2013 and 2014. Experimental results demonstrate that the proposed approach significantly outperforms the state of the art methods.

• Journal of the Korea Society for Simulation
• /
• v.19 no.1
• /
• pp.103-111
• /
• 2010

Syntactic analysis is an important step in natural language processing. However, we cannot use the syntactic analyzer in Korean for low performance and without robustness. We propose new robust, high speed and high performance Korean syntactic analyzer using CRFs. We treat a parsing problem as a labeling problem. We use a cascaded chunking for Korean parsing. We label syntactic information to each Eojeol at each step using CRFs. CRFs use part-of-speech tag and Eojeol syntactic tag features. Our experimental results using 10-fold cross validation show significant improvement in the robustness, speed and performance of long Korea sentences.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.2
• /
• pp.381-394
• /
• 2016

Graphical model represents conditional dependencies between variables as a graph with nodes and edges. It is widely used in various fields including physics, economics, and biology to describe complex association. Conditional dependencies can be estimated from a inverse covariance matrix, where zero off-diagonal elements denote conditional independence of corresponding variables. This paper proposes a efficient BCDR (block coordinate descent with random permutation) algorithm using graphics processing units and random permutation for the CONCORD (convex correlation selection method) based on the BCD (block coordinate descent) algorithm, which estimates a inverse covariance matrix based on pseudo-likelihood. We conduct numerical studies for two network structures to demonstrate the efficiency of the proposed algorithm for the CONCORD in terms of computation times.

• ETRI Journal
• /
• v.42 no.1
• /
• pp.90-100
• /
• 2020

Named entity recognition (NER) continues to be an important task in natural language processing because it is featured as a subtask and/or subproblem in information extraction and machine translation. In Urdu language processing, it is a very difficult task. This paper proposes various deep recurrent neural network (DRNN) learning models with word embedding. Experimental results demonstrate that they improve upon current state-of-the-art NER approaches for Urdu. The DRRN models evaluated include forward and bidirectional extensions of the long short-term memory and back propagation through time approaches. The proposed models consider both language-dependent features, such as part-of-speech tags, and language-independent features, such as the "context windows" of words. The effectiveness of the DRNN models with word embedding for NER in Urdu is demonstrated using three datasets. The results reveal that the proposed approach significantly outperforms previous conditional random field and artificial neural network approaches. The best f-measure values achieved on the three benchmark datasets using the proposed deep learning approaches are 81.1%, 79.94%, and 63.21%, respectively.

• Journal of Internet Computing and Services
• /
• v.17 no.5
• /
• pp.111-119
• /
• 2016

In recent years, the opinion analysis is one of the key research fronts of any domain. Opinion target extraction is an essential process of opinion analysis. Target is usually referred to noun or noun phrase in an entity which is deliberated by the opinion holder. Extraction of opinion target facilitates the opinion analysis more precisely and in addition helps to identify the opinion polarity i.e. users can perceive opinion in detail of a target including all its features. One of the most commonly employed algorithms is a sequence labeling algorithm also called Conditional Random Fields. In present article, recent opinion target extraction approaches are reviewed based on sequence labeling algorithm and it features combinations by analyzing and comparing these approaches. The good selection of features combinations will in some way give a good or better accuracy result. Features combinations are an essential process that can be used to identify and remove unneeded, irrelevant and redundant attributes from data that do not contribute to the accuracy of a predictive model or may in fact decrease the accuracy of the model. Hence, in general this review eventually leads to the contribution for the opinion analysis approach and assist researcher for the opinion target extraction in particular.

• Proceedings of the Korean Society for Language and Information Conference
• /
• 2007.11a
• /
• pp.303-310
• /
• 2007

Named Entity Recognition (NER) is always limited by its lower recall resulting from the asymmetric data distribution where the NONE class dominates the entity classes. This paper presents an approach that exploits non-local information to improve the NER recall. Several kinds of non-local features encoding entity token occurrence, entity boundary and entity class are explored under Conditional Random Fields (CRFs) framework. Experiments on SIGHAN 2006 MSRA (CityU) corpus indicate that non-local features can effectively enhance the recall of the state-of-the-art NER systems. Incorporating the non-local features into the NER systems using local features alone, our best system achieves a 23.56% (25.26%) relative error reduction on the recall and 17.10% (11.36%) relative error reduction on the F1 score; the improved F1 score 89.38% (90.09%) is significantly superior to the best NER system with F1 of 86.51% (89.03%) participated in the closed track.

• Smart Media Journal
• /
• v.10 no.2
• /
• pp.55-60
• /
• 2021

In this research, we design and perform experiment to investigate whether neuronal activity patterns elicited while solving game tasks are spontaneously reactivated in during sleep. In order to recognize human activity EEG-fMRI signals are used at the same time. Experimental results shows that reward for the success of tasks performed before sleeping have an effect on sleep brain activity. The study uncovers a neural mechanism whereby rewarded life experiences are preferentially replayed and consolidated while we sleep.

• Journal of KIISE
• /
• v.43 no.6
• /
• pp.678-685
• /
• 2016

Named Entity Recognition and Classification (NERC) is a task for recognition and classification of named entities such as a person's name, location, and organization. There have been various studies carried out on Korean NERC, but they have some problems, for example lacking some features as compared with English NERC. In this paper, we propose a method that uses word embedding as features for Korean NERC. We generate a word vector using a Continuous-Bag-of-Word (CBOW) model from POS-tagged corpus, and a word cluster symbol using a K-means algorithm from a word vector. We use the word vector and word cluster symbol as word embedding features in Conditional Random Fields (CRFs). From the result of the experiment, performance improved 1.17%, 0.61% and 1.19% respectively for TV domain, Sports domain and IT domain over the baseline system. Showing better performance than other NERC systems, we demonstrate the effectiveness and efficiency of the proposed method.

• Journal of KIISE:Software and Applications
• /
• v.37 no.11
• /
• pp.837-844
• /
• 2010

Prediction of the prosodic phrase boundary is one of the most important natural language processing tasks. We propose, for the natural prediction of the Korean prosodic phrase boundary, a statistical approach incorporating efficient learning features. These new features reflect the factors that affect generation of the prosodic phrase boundary better than existing learning features. Notably, moreover, such learning features, extracted according to the hand-crafted prosodic phrase boundary prediction rule, impart higher accuracy. We developed a statistical model for Korean prosodic phrase boundaries based on the proposed new features. The results were 86.63% accuracy for three levels (major break, minor break, no break) and 81.14% accuracy for six levels (major break with falling tone/rising tone, minor break with falling tone/rising tone/middle tone, no break).