DOI QR코드

DOI QR Code

Exploring Feature Selection Methods for Effective Emotion Mining

효과적 이모션마이닝을 위한 속성선택 방법에 관한 연구

  • Eo, Kyun Sun (KK Business School, Sungkyunkwan University) ;
  • Lee, Kun Chang (Global Business Administration/Dept. of Health Sciences & & Technology, SAIHST Sungkyunkwan University)
  • 어균선 (성균관대학교 경영대학) ;
  • 이건창 (성균관대학교 글로벌경영학과/삼성융합의과학원 융합의과학과)
  • Received : 2018.11.28
  • Accepted : 2019.03.20
  • Published : 2019.03.28

Abstract

In the era of SNS, many people relies on it to express their emotions about various kinds of products and services. Therefore, for the companies eagerly seeking to investigate how their products and services are perceived in the market, emotion mining tasks using dataset from SNSs become important much more than ever. Basically, emotion mining is a branch of sentiment analysis which is based on BOW (bag-of-words) and TF-IDF. However, there are few studies on the emotion mining which adopt feature selection (FS) methods to look for optimal set of features ensuring better results. In this sense, this study aims to propose FS methods to conduct emotion mining tasks more effectively with better outcomes. This study uses Twitter and SemEval2007 dataset for the sake of emotion mining experiments. We applied three FS methods such as CFS (Correlation based FS), IG (Information Gain), and ReliefF. Emotion mining results were obtained from applying the selected features to nine classifiers. When applying DT (decision tree) to Tweet dataset, accuracy increases with CFS, IG, and ReliefF methods. When applying LR (logistic regression) to SemEval2007 dataset, accuracy increases with ReliefF method.

블로그, 소셜 미디어 등의 발달로 인해 점점 더 많은 사람들이 본인의 의견이나 감정을 표현하기 위해 온라인상에서 텍스트 문장을 작성한다. 그리고 이같은 온라인 텍스트 문장속에 숨겨져 있는 긍정 또는 부정등의 감성을 찾아내는 연구분야를 감성분석 이라고 한다. 그중에서도 이모션 마이닝은 사람들의 구체적인 이모션을 찾아내는데 초점을 맞춘 연구분야이다. 본 연구에서는 속성선택 방법과 단일 및 앙상블 분류기를 조합하여 효과적인 이모션 마이닝 예측모델을 제시하고자 한다. 이를 위해 두가지 대표적인 오픈 데이터인 Tweet와 SemEval2007 데이터를 이용하여 TF-IDF를 계산하고 백 오브 워즈(BOW: bag-of-words) 형태로 속성 셋을 구성하였다. 그리고 효과적인 이모션 마이닝이 될 수 있는 최적의 속성을 선택하기 위하여 상관관계 기반 속성선택(CFS), 정보획득 속성선택 (IG), 그리고 ReliefF 등 세가지 속성선택 방법을 적용하였다. 선택된 속성을 이용하여 아홉가지 분류기 모델로 이모션 마이닝의 정확도를 비교하였다. 실험 결과, Tweet 데이터는 의사결정나무(DT)가 CFS, IG, ReliefF에 의한 속성을 이용할 경우 정확도가 상승했고, 랜덤서브스페이스(RS)는 CFS, IG에 선택된 속성을 사용할 경우 정확도가 상승했다. SemEval2007 데이터는 ReliefF에 의해 선택된 속성으로 로지스틱 회귀분석(LR)을 적용하였을 때 정확도가 상승했고, 나이브 베이지안 네트워크(NBN)은 CFS, IG에 의한 속성을 사용할 경우 정확도가 상승하였다.

Keywords

DJTJBT_2019_v17n3_107_f0001.png 이미지

Fig. 1. Procedure of this study

Table 1. Emotion mining studies

DJTJBT_2019_v17n3_107_t0001.png 이미지

Table 2. Confusion matrix

DJTJBT_2019_v17n3_107_t0002.png 이미지

Table 3. The number of selected features

DJTJBT_2019_v17n3_107_t0003.png 이미지

Table 4. Tweet results

DJTJBT_2019_v17n3_107_t0004.png 이미지

Table 5. SemEval 2007 results

DJTJBT_2019_v17n3_107_t0005.png 이미지

Table 6. T-test Result of Accuracy

DJTJBT_2019_v17n3_107_t0006.png 이미지

References

  1. J. A. Balazs & J. D. Velasquez. (2016). Opinion mining and information fusion: a survey. Information Fusion, 27, 95-110. https://doi.org/10.1016/j.inffus.2015.06.002
  2. H. L. Yang & Q. F. Lin. (2018). Opinion mining for multiple types of emotion-embedded products/services through evolutionary strategy. Expert Systems with Applications, 99, 44-55. https://doi.org/10.1016/j.eswa.2018.01.022
  3. M. V. Mantyla, D. Graziotin & M. Kuutila. (2018). The evolution of sentiment analysis-A review of research topics, venues, and top cited papers. Computer Science Review, 27, 16-32. https://doi.org/10.1016/j.cosrev.2017.10.002
  4. Y. Liu, J. W. Bi & Z. P. Fan. (2017). Multi-class sentiment classification: The experimental comparisons of feature selection and machine learning algorithms. Expert Systems with Applications, 80, 323-339. https://doi.org/10.1016/j.eswa.2017.03.042
  5. T. Danisman & A. Alpkocak. (2008, April). Feeler: Emotion classification of text using vector space model. In AISB 2008 Convention Communication, Interaction and Social Intelligence (Vol. 1, p. 53).
  6. C., Strapparava & R. Mihalcea. (2007). Semeval-2007 task 14: Affective text. In Proceedings of the 4th international workshop on semantic evaluations (pp. 70-74). Association for Computational Linguistics.
  7. N. Gupta, M. Gilbert & G. D. Fabbrizio. (2013). Emotion detection in email customer care. Computational Intelligence, 29(3), 489-505. https://doi.org/10.1111/j.1467-8640.2012.00454.x
  8. M. Hasan, E. Agu & E. Rundensteiner. (2014). Using hashtags as labels for supervised learning of emotions in twitter messages. In ACM SIGKDD Workshop on Health Informatics, New York, USA.
  9. C. Quan & F. Ren. (2016). Weighted high-order hidden Markov models for compound emotions recognition in text. Information Sciences, 329, 581-596. https://doi.org/10.1016/j.ins.2015.09.050
  10. M. A. Hall. (1999). Correlation-based feature selection for machine learning.
  11. M. Robnik-Sikonja & I. Kononenko. (2003). Theoretical and empirical analysis of ReliefF and RReliefF. Machine learning, 53(1-2), 23-69. https://doi.org/10.1023/A:1025667309714
  12. D. R. Cox. (1958). The regression analysis of binary sequences. Journal of the Royal Statistical Society. Series B (Methodological), 215-242.
  13. S. K. Murthy. (1998). Automatic construction of decision trees from data: A multi-disciplinary survey. Data mining and knowledge discovery, 2(4), 345-389. https://doi.org/10.1023/A:1009744630224
  14. E. C. Bae & K. C. Lee. (2016). Predicting Stock Liquidity by Using Ensemble Data Mining Methods", Journal of The Korea Society of computer and Information, 21(6), 9-19, https://doi.org/10.9708/JKSCI.2016.21.6.009
  15. S. Park, K. M. Yang & S. B. Cho. (2013). A Hierarchical CPV Solar Generation Tracking System based on Modular Bayesian Network. Journal of KIISE: Software and Applications, 41.
  16. V. Vapnik. (2013). The nature of statistical learning theory. Springer science & business media.
  17. M. H. Song, J. Lee, S. P. Cho & K. J. Lee. (2005). SVM Classifier for the Detection of Ventricular Fibrillation, The Institute of Electronics Engineers of Korea - System and Control, 42(5), 27-34.
  18. M. Ballings, D. Van den Poel, N. Hespeels & R. Gryp. (2015). Evaluating multiple classifiers for stock price direction prediction. Expert Systems with Applications, 42(20), 7046-7056. https://doi.org/10.1016/j.eswa.2015.05.013
  19. D. H. Wolpert. (1992). Stacked generalization. Neural networks, 5(2), 241-259. https://doi.org/10.1016/S0893-6080(05)80023-1
  20. J. H. Lee & J. G. Baek. (2018). RTC(Real-Time Contrast) Control Chart using Random Forest based Multi-Class Classifier, Journal of the Korean Institute of Industrial Engineers, 44(4), 306-315. https://doi.org/10.7232/JKIIE.2018.44.4.306
  21. T. K. Ho. (1998). The Random Subspace Method for Constructing Decision Forests, IEEE Trans. Pattern Analysis and Machine Intelligence, 20(8), 832-844. https://doi.org/10.1109/34.709601
  22. W. Wang, L. Chen, K. Thirunarayan & A. P. Sheth. (2012). Harnessing twitter big data for automatic emotion identification. In Privacy, Security, Risk and Trust (PASSAT), 2012 International Conference on and 2012 International Confernece on Social Computing (SocialCom), IEEE, 587-592.
  23. A. Yadollahi, A. G. Shahraki & O. R. Zaiane. (2017). Current state of text sentiment analysis from opinion to emotion mining. ACM Computing Surveys (CSUR), 50(2), 25.
  24. S. Arlot & A. Celisse. (2010). A survey of cross-validation procedures for model selection. Statistics surveys, 4, 40-79. https://doi.org/10.1214/09-SS054