• Journal of the Korea Society of Computer and Information
• /
• v.26 no.8
• /
• pp.55-63
• /
• 2021

Since social big data often includes new words or proper nouns, statistical morphological analysis methods have been widely used to process them properly which are based on the frequency of occurrence of each word. However, these methods do not properly recognize compound nouns, and thus have a problem in that the accuracy of keyword extraction is lowered. This paper presents a method to extract compound nouns in keyword analysis of social big data. The proposed method creates a candidate group of compound nouns by combining the words obtained through the morphological analysis step, and extracts compound nouns by examining their frequency of appearance in a given review. Two algorithms have been proposed according to the method of constructing the candidate group, and the performance of each algorithm is expressed and compared with formulas. The comparison result is verified through experiments on real data collected online, where the results also show that the proposed method is suitable for real-time processing.

• ETRI Journal
• /
• v.39 no.5
• /
• pp.652-661
• /
• 2017

A mention has a noun or noun phrase as its head and constructs a chunk that defines any meaning, including a modifier. Mention detection refers to the extraction of mentions from a document. In mentions, coreference resolution refers to determining any mentions that have the same meaning. Pointer networks, which are models based on a recurrent neural network encoder-decoder, outputs a list of elements corresponding to an input sequence. In this paper, we propose mention detection using pointer networks. This approach can solve the problem of overlapped mention detection, which cannot be solved by a sequence labeling approach. The experimental results show that the performance of the proposed mention detection approach is F1 of 80.75%, which is 8% higher than rule-based mention detection, and the performance of the coreference resolution has a CoNLL F1 of 56.67% (mention boundary), which is 7.68% higher than coreference resolution using rule-based mention detection.

• Annual Conference on Human and Language Technology
• /
• 2000.10d
• /
• pp.19-25
• /
• 2000

명사 추출기는 정보검색, 문서분류, 문서요약, 정보추출 등의 분야에서 사용되고 있으며, 정확한 명사 추출과 빠른 색인 속도는 이들 시스템 성능과 밀접한 관계가 있다. 한국어에서 명사를 추출하기 위해서는 형태소 분석이 필요한데, 본 논문에서는 대량의 품사부착된 말뭉치로부터 추출한 분석배제 정보와 후절어를 이용함으로써 형태소 분석을 생략하거나 보다 단순한 처리에 의해 명사를 추출하는 방법을 제안한다. 또한 형태소 분석시 복잡한 음운 현상을 처리하기 위해 많은 음운 규칙을 적용하는 대신 음운 복원 정보를 사용하여 음운 현상을 처리하는 방법을 제안한다. 실험결과에 의하면, 제안된 방법에 의한 명사추출기는 비교적 높은 정확률과 재현율을 나타내며, 빠른 속도를 보였다.

• Proceedings of the Korean Society for Cognitive Science Conference
• /
• 2000.06a
• /
• pp.309-314
• /
• 2000

본 논문에서는 양국어로 된 병렬 코퍼스로부터 복합어의 역어를 추출하기 위한 정렬 방법을 제시한다. 여기에서는 개념어에 대한 양국어 공기정보를 사용하여 기본 정렬을 하고, 인접한 개념어로 정렬의 단위를 확장했다. 또한 재추정 기법을 사용하여 대역 확률을 계산함으로써 보다 높은 정확률을 얻을 수 있었다. 본 논문에서 제안한 방법을 적용하여 139,265개의 영어 어절로 이루어진 우루과이 라운드 영한 병렬 코퍼스에 대해서 실험한 결과 2,290개의 대역어쌍을 얻었고, 그 정확률은 74%였다.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.10b
• /
• pp.184-186
• /
• 2001

본 논문은 고유명사를 활용하여 특정 정보를 좀더 효율적으로 추출하기 위한 연구이며, Named Entity의 한 범주인 사람 이름에 대하여 어휘 사전이나 실마리 사전의 사용 없이 초기에 주어지는 몇 개의 인칭 명사들을 태그가 부착되지 않은 코퍼스에 적용시켜 고유명사 추출을 위한 패턴을 학습하고, 그 패턴을 적용하여 새로운 고유명사를 생성해 내는 작업을 통해 인칭 명사들을 효율적으로 추출할 수 있는 방법을 제안한다.

• Journal of Advanced Navigation Technology
• /
• v.20 no.6
• /
• pp.617-624
• /
• 2016

The hangul word embedding should be performed certainly process for noun extraction. Otherwise, it should be trained words that are not necessary, and it can not be derived efficient embedding results. In this paper, we propose model that can retrieve more efficiently by query language expansion using hangul word embedded, apriori, and text mining. The word embedding and apriori is a step expanding query language by extracting association words according to meaning and context for query language. The hangul text mining is a step of extracting similar answer and responding to the user using noun extraction, TF-IDF, and cosine similarity. The proposed model can improve accuracy of answer by learning the answer of specific domain and expanding high correlation query language. As future research, it needs to extract more correlation query language by analysis of user queries stored in database.

• Journal of Information Processing Systems
• /
• v.3 no.1
• /
• pp.33-37
• /
• 2007

Semantic clustering is important to various fields in the modem information society. In this work we applied the Independent Component Analysis method to the extraction of the features of latent concepts. We used verb and object noun information and formulated a concept as a linear combination of verbs. The proposed method is shown to be suitable for our framework and it performs better than a hierarchical clustering in latent semantic space for finding out invisible information from the data.

• The Transactions of the Korea Information Processing Society
• /
• v.1 no.4
• /
• pp.461-468
• /
• 1994

This paper describes the automatic extraction of collocations based on corpus. The collocations are extracted from corpus using cooccurrence frequency and mutual information between words. In English, 5 types of collocations are defined. These collocations are transitive verb and object, intransitive verb and subject, adjective and noun, verb and adverb, and adverb and adjective. In this paper another type of collocation is recognized and extracted, which consists of verb and preposition. So 6 types of collocations are extracted based on corpus.

• Journal of the Korean Society for information Management
• /
• v.36 no.4
• /
• pp.7-19
• /
• 2019

The aim of this study is to develop a technique for keyword extraction in Korean-Chinese text in the modern period. We considered a Korean morphological analyzer and a particle in classical Chinese as a possible method for this study. We applied our method to the journal "Sisachongbo," employing proper-noun dictionaries and a list of stop words to extract index terms. The results show that our system achieved better performance than a Chinese morphological analyzer in terms of recall and precision. This study is the first research to develop an automatic indexing system in the traditional Korean-Chinese mixed text.

• Journal of Information Management
• /
• v.39 no.3
• /
• pp.163-177
• /
• 2008

In this paper, we analyzed the relationship between user query keywords that is used to search NDSL and technical terms extracted from NDSL journals. For the analysis, we extracted about 833,000 query keywords from NDSL search logs during nearly 17 months and approximately 41,000,000 technical terms from NDSL, INSPEC, FSTA journals. And we used only the English noun phrase in extracted those and then we did an experiment on analysis of equality, relationship analysis and frequency analysis.