• Annual Conference on Human and Language Technology
• /
• 2020.10a
• /
• pp.563-568
• /
• 2020

본 연구에서는 블랙박스로 알려진 딥러닝 모델에 설명 근거를 제공할 수 있는 설명자 모델을 적용해 보았다. 영화평 감성 분석을 위해 MLP, CNN으로 구성된 딥러닝 모델과 결정트리의 앙상블인 Gradient Boosting 모델을 이용하여 감성 분류기를 구축하였다. 설명자 모델로는 기울기(gradient)을 기반으로 하는 IG와 레이어 사이의 가중치(weight)을 기반으로 하는 CAM, 그리고 설명가능한 대리 모델을 이용하는 LIME과 입력 속성에 대한 선형모델을 추정하는 SHAP을 사용하였다. 설명자 모델의 특성을 보기 위하여 히트맵과 관련성 높은 N개의 속성을 추출해 보았다. 설명자가 제공하는 기여도에 따라 입력 속성을 제거해 가며 분류기 성능 변화를 측정하는 정량적 평가도 수행하였다. 또한, 사람의 판단 근거와의 일치도를 살펴볼 수 있는 '설명 근거 정확도'라는 새로운 평가 방법을 제안하여 적용해 보았다.

• Science of Emotion and Sensibility
• /
• v.17 no.1
• /
• pp.71-78
• /
• 2014

We propose the sentiment pattern as a novel sentiment feature for more accurate text sentiment analysis, and introduce the rating inference of movie reviews using it. The text sentiment analysis is a task that recognizes and classifies sentiment of text whether it is positive or negative. For that purpose, the sentiment feature is used, which includes sentiment words and phrase pattern that have specific sentiment like positive or negative. The previous researches for the sentiment analysis, however, have a limit to understand accurately total sentiment of either a sentence or text because they consider the sentiment of sentiment words and phrase patterns independently. Therefore, we propose the sentiment pattern that is defined by arranging semantically all sentiment in a sentence, and use them as a new sentiment feature for the rating inference that is one of the detail subjects of the sentiment analysis. In order to verify the effect of proposed sentiment pattern, we conducted experiments of rating inference. Ratings of test reviews is inferred by using a probabilistic method with sentiment features including sentiment patterns extracted from training reviews. As a result, it is shown that the result of rating inference with sentiment patterns are more accurate than that without sentiment patterns.

• KIPS Transactions on Software and Data Engineering
• /
• v.4 no.3
• /
• pp.121-128
• /
• 2015

A given product's online product reviews build up to form largely positive or negative reviews or mixed reviews that include both the positive and negative reviews. While the homogeneously positive or negative reviews help readers identify the generally praised or criticized product, the mixed reviews with minority opinions potentially contain valuable information about the product. We present a method of retrieving minority opinions from the online product reviews using the skewness of positive/negative reviews. The proposed method first classifies the positive/negative product reviews using a sentiment dictionary and then calculates the skewness of the classified results to identify minority reviews. Minority review retrieval experiments were conducted on smartphone and movie reviews, and the F1-measures were 24.6% (smartphone) and 15.9% (movie) and the accuracies were 56.8% and 46.8% when the individual reviews' sentiment classification accuracies were 85.3% and 78.8%. The theoretical performance of minority review retrieval is also discussed.

• Annual Conference on Human and Language Technology
• /
• 2009.10a
• /
• pp.56-60
• /
• 2009

오늘날 인터넷은 개인의 감정, 의견을 서로 공유할 수 있는 공간이 되고 있다. 하지만 인터넷에는 너무나 방대한 문서가 존재하기 때문에 다른 사용자들의 감정, 의견 정보를 개인의 의사 결정에 활용하기가 쉽지 않다. 최근 들어 감정이나 의견을 자동으로 추출하기 위한 연구가 활발하게 진행되고 있으며, 감정 분석에 관한 기존 연구들은 대부분 어구의 극성(polarity) 정보가 있는 감정 사전을 사용하고 있다. 하지만 인터넷에는 나날이 신조어가 새로 생기고 언어 파괴 현상이 자주 일어나기 때문에 사전에 기반한 방법은 한계가 있다. 본 논문은 감정 분석 문제를 긍정과 부정으로 구분하는 이진 분류 문제로 본다. 이진 분류 문제에서 탁월한 성능을 보이는 Support Vector Machines(SVM)을 사용하며, 문서들 간의 유사도 계산을 위해 문장의 부분 문자열을 비교하는 문자열 커널을 사용한다. 실험 결과, 실제 영화평에서 제안된 모델이 비교 대상으로 삼은 Bag of Words(BOW) 모델보다 안정적인 성능을 보였다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2018.05a
• /
• pp.358-361
• /
• 2018

비정형 데이터 증가로 텍스트 마이닝을 사용해 데이터를 분석하는 연구가 주목받고 있다. 감성분석은 단어와 문맥을 분석하여 텍스트의 감정을 파악하는 기술이다. 본 논문에서는 감성분석 연구 동향, 적용분야, 방법론에 관해 분석하고 기술하려 한다. 감성분석은 2001년 채팅의 감정을 분석하면서 시작되었고, 2008년부터 본격적으로 연구가 진행되었다. 감성분석은 SNS, 상품 후기, 영화평, 뉴스 기사 등 다양한 데이터에 적용되고 있으며, 사회이슈 찬반 분석과 장소 선호도 분석 등 다양한 연구에서 사용되었다. 감성분석 방법은 감성사전을 이용하는 방식과 기계학습을 사용하는 방식으로 나누어지며 분석 방법을 발전시키기 위한 연구가 진행되고 있다.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.24 no.3
• /
• pp.292-297
• /
• 2014

As a Social Network Service (SNS) has become an integral part of our everyday lives, millions of users can express their opinion and share information regardless of time and place. Hence sentiment analysis using micro-blogs has been studied in various field to know people's opinion on particular topics. Most of previous researches on movie reviews consider only positive and negative sentiment and use it to predict movie rating. As people feel not only positive and negative but also various emotion, the sentiment that people feel while watching a movie need to be classified in more detail to extract more information than personal preference. We measure sentiment distributions of each movie from tweets according to the Thayer's model. Then, we find similar movies by calculating similarity between each sentiment distributions. Through the experiments, we verify that our method using micro-blogs performs better than using only genre information of movies.

• Science of Emotion and Sensibility
• /
• v.18 no.1
• /
• pp.103-110
• /
• 2015

Sentiment lexicon is an essential element for expressing sentiment on a text or recognizing sentiment from a text. We propose a graph-based semi-supervised learning method to construct a sentiment dictionary as sentiment lexicon set. In particular, we focus on the construction of domain-specific sentiment dictionary. The proposed method makes up a graph according to lexicons and proximity among lexicons, and sentiments of some lexicons which already know their sentiment values are propagated throughout all of the lexicons on the graph. There are two typical types of the sentiment lexicon, sentiment words and sentiment phrase, and we construct a sentiment dictionary by creating each graph of them and infer sentiment of all sentiment lexicons. In order to verify our proposed method, we constructed a sentiment dictionary specific to the movie domain, and conducted sentiment classification experiments with it. As a result, it have been shown that the classification performance using the sentiment dictionary is better than the other using typical general-purpose sentiment dictionary.

• Korean Journal of Cognitive Science
• /
• v.21 no.4
• /
• pp.669-696
• /
• 2010

Assuming that the whole meaning of a document is a composition of the meanings of each part, this paper proposes to study the automatic grading of movie reviews which contain sentimental expressions. This will be accomplished by calculating the values of semantic segments and performing data classification for each review. The ARSSA(The Automatic Rating System for Sentiment analysis using an Appraisal dictionary) system is an effort to model decision making processes in a manner similar to that of the human mind. This aims to resolve the discontinuity between the numerical ranking and textual rationalization present in the binary structure of the current review rating system: {rate: review}. This model can be realized by performing analysis on the abstract menas extracted from each review. The performance of this system was experimentally calculated by performing a 10-fold Cross-Validation test of 1000 reviews obtained from the Naver Movie site. The system achieved an 85% F1 Score when compared to predefined values using a predefined appraisal dictionary.

• 주택과사람들
• /
• s.189
• /
• pp.100-107
• /
• 2006

600여 만 평에 걸쳐진 대관령 능선을 따라 아름다운 설원이 펼쳐지는 평창. 이곳을 찾는 사람들은 누구나 영화와 드라마의 잔상을 떠올리며 겨울 동화의 주인공이 되어 돌아간다. 이처럼 설원의 도시, 고원 휴양지로 알려진 평창이 2014년 동계 올림픽을 준비하며 야심찬 계획을 세우고 있다. 과연 평창에는 어떤 움직임이 일어나고 있는지 사뭇 궁금해진다. 강원도 평창, 그 현장 속으로···.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.17 no.3
• /
• pp.167-173
• /
• 2017

With increased interest in the fourth industrial revolution represented by artificial intelligence, it has been very active to utilize bigdata and machine learning techniques in almost areas of society. Also, such activities have been realized by development of forecasting systems in various applications. Especially in the movie industry, there have been numerous attempts to predict whether they would be success or not. In the past, most of studies considered only the static factors in the process of prediction, but recently, several efforts are tried to utilize realtime social bigdata produced in SNS. In this paper, we propose the prediction technique utilizing various feedback information such as news articles, blogs and reviews as well as static factors of movies. Additionally, we also experimentally evaluate whether the proposed technique could precisely forecast their revenue targeting on the relatively successful movies.