Word-Level Embedding to Improve Performance of Representative Spatio-temporal Document Classification
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Tokenization is the process of segmenting the input text into smaller units of text, and it is a preprocessing taskthat is mainly performed to improve the efficiency of the machine learning process. Various tokenizationmethods have been proposed for application in the field of natural language processing, but studies haveprimarily focused on efficiently segmenting text. Few studies have been conducted on the Korean language toexplore what tokenization methods are suitable for document classification task. In this paper, an exploratorystudy was performed to find the most suitable tokenization method to improve the performance of arepresentative spatio-temporal document classifier in Korean. For the experiment, a convolutional neuralnetwork model was used, and for the final performance comparison, tasks were selected for documentclassification where performance largely depends on the tokenization method. As a tokenization method forcomparative experiments, commonly used Jamo, Character, and Word units were adopted. As a result of theexperiment, it was confirmed that the tokenization of word units showed excellent performance in the case ofrepresentative spatio-temporal document classification task where the semantic embedding ability of the tokenitself is important.
Abstract
Tokenization is the process of segmenting the input text into smaller units of text, and it is a preprocessing taskthat is mainly performed to improve the efficiency of the machine learning process. Various tokenizationmethods have been proposed for application in the field of natural language processing, but studies haveprimarily focused on efficiently segmenting text. Few studies have been conducted on the Korean language toexplore what tokenization methods are suitable for document classification task. In this paper, an exploratorystudy was performed to find the most suitable tokenization method to improve the performance of arepresentative spatio-temporal document classifier in Korean. For the experiment, a convolutional neuralnetwork model was used, and for the final performance comparison, tasks were selected for documentclassification where performance largely depends on the tokenization method. As a tokenization method forcomparative experiments, commonly used Jamo, Character, and Word units were adopted. As a result of theexperiment, it was confirmed that the tokenization of word units showed excellent performance in the case ofrepresentative spatio-temporal document classification task where the semantic embedding ability of the tokenitself is important.
참고문헌 (28)
- A. Mucherino, "Data Mining in Agriculture", pp.83~106 (2009), Springer
- Y. Wang, "Classification of web documents using a naive bayes method,”in Proceedings", 15th IEEE International Conference on Tools with Artificial Intelligence, pp.560~564 (2003)
- S. Mayor, "Document classification using support vector machine", International Journal of Engineering Science and Technology, 4(4), pp.1741~1745 (2012)
- Wan M.U. Noormanshah, "Document Categorization Using Decision Tree: Preliminary Study", International Journal of Engineering & Technology, 7(4.34), pp.437~ (2018), Science Publishing Corporation[DOI]
- "Detecting and extracting events from text documents" (2016)[URL]
- A. Badia, "Text extraction of spatial and temporal information", Proceedings of 2007 IEEE Intelligence and Security Informatics, pp.381~381 (2007)
- 임채규, "Survey of Temporal Information Extraction", JIPS(Journal of Information Processing Systems), 15(4), pp.931~956 (2019), 한국정보처리학회[DOI]
- Abdelkoui Feriel, "Automatic Extraction of Spatio-Temporal Information from Arabic Text Documents", International Journal of Computer Science and Information Technology, 7(5), pp.97~107 (2015), Academy and Industry Research Collaboration Center (AIRCC)[DOI]
- Byoungwook Kim, "A Convolution Neural Network-Based Representative Spatio-Temporal Documents Classification for Big Text Data", Applied Sciences, 12(8), pp.3843~ (2022), MDPI AG[DOI]
- Jingnian Chen, "Feature selection for text classification with Naïve Bayes", Expert Systems with Applications, 36(3), pp.5432~5435 (2009), Elsevier BV[DOI]
- "How to build and apply Naive Bayes classification for spam filtering" (2020)[URL]
- Vikramjit Mitra, "Text classification: A least square support vector machine approach", Applied Soft Computing, 7(3), pp.908~914 (2007), Elsevier BV[DOI]
- M. Z. Islam, "A semantics aware random forest for text classification", Proceedings of the 28th ACM International Conference on Information and Knowledge Management, pp.1061~1070 (2019)
- Y. LeCun, "Deep learning", Nature, 521(7553), pp.436~444 (2015)
- Z. Zhong, "Computer Vision–ECCV 2020", pp.191~207 (2020), Springer
- "Efficient estimation of word representations in vector space" (2013)[URL]
- T. Huang, "A CNN model for SMS spam detection", Proceedings of 2019 4th International Conference on Mechanical, Control and Computer Engineering (ICMCCE), pp.851~85110 (2019)
- Sisi Liu, "Sequence encoding incorporated CNN model for Email document sentiment classification", Applied Soft Computing, 102, pp.107104~ (2021), Elsevier BV[DOI]
- Emmanuel Mutabazi, "A Review on Medical Textual Question Answering Systems Based on Deep Learning Approaches", Applied Sciences, 11(12), pp.5456~ (2021), MDPI AG[DOI]
- Mira Kim, "Automated Classification of Online Sources for Infectious Disease Occurrences Using Machine-Learning-Based Natural Language Processing Approaches", International Journal of Environmental Research and Public Health, 17(24), pp.9467~ (2020), MDPI AG[DOI]
- B. Ban, "A survey on awesome Korean NLP datasets", Proceedings of 2022 13th International Conference on Information and Communication Technology Convergence (ICTC), pp.1615~1620 (2022)
- Dang Thanh Vu, "Text Data Augmentation for the Korean Language", Applied Sciences, 12(7), pp.3425~ (2022), MDPI AG[DOI]
- Yoojoong Kim, "A pre-trained BERT for Korean medical natural language processing", Scientific Reports, 12(1) (2022), Springer Science and Business Media LLC[DOI]
- J. Shin, "Constructing Korean abusive language dataset using machine translation", Proceedings of the Korea Computer Congress (KCC) (2022)
- Jaejin Seo, "TA-SBERT: Token Attention Sentence-BERT for Improving Sentence Representation", IEEE Access, 10, pp.39119~39128 (2022), Institute of Electrical and Electronics Engineers (IEEE)[DOI]
- Cagri Toraman, "Impact of Tokenization on Language Models: An Analysis for Turkish", ACM Transactions on Asian and Low-Resource Language Information Processing, 22(4), pp.1~21 (2023), Association for Computing Machinery (ACM)[DOI]
- M. Alkaoud, "On the importance of tokenization in Arabic embedding models", Proceedings of the 5th Arabic Natural Language Processing Workshop, pp.119~129 (2020)
- Shaobo Li, "DeepPatent: patent classification with convolutional neural networks and word embedding", Scientometrics, 117(2), pp.721~744 (2018), Springer Science and Business Media LLC[DOI]