• The Transactions of the Korea Information Processing Society
• /
• v.13 no.4
• /
• pp.174-180
• /
• 2024

In this paper, we explore an emotion classification method through multimodal learning utilizing wav2vec 2.0 and KcELECTRA models. It is known that multimodal learning, which leverages both speech and text data, can significantly enhance emotion classification performance compared to methods that solely rely on speech data. Our study conducts a comparative analysis of BERT and its derivative models, known for their superior performance in the field of natural language processing, to select the optimal model for effective feature extraction from text data for use as the text processing model. The results confirm that the KcELECTRA model exhibits outstanding performance in emotion classification tasks. Furthermore, experiments using datasets made available by AI-Hub demonstrate that the inclusion of text data enables achieving superior performance with less data than when using speech data alone. The experiments show that the use of the KcELECTRA model achieved the highest accuracy of 96.57%. This indicates that multimodal learning can offer meaningful performance improvements in complex natural language processing tasks such as emotion classification.

• Journal of Korea Multimedia Society
• /
• v.13 no.12
• /
• pp.1786-1797
• /
• 2010

Automatic understanding of contents in document image is a very hard problem due to involvement with mathematically challenging problems originated mainly from the over-determined system induced by document segmentation process. In both academic and industrial areas, there have been incessant and various efforts to improve core parts of content retrieval technologies by the means of separating out segmentation related issues using semi-structured document, e.g., invoice,. In this paper we proposed classification models for text lines on invoice document in which text lines were clustered into the five categories in accordance with their contents: purchase order header, invoice header, summary header, surcharge header, purchase items. Our investigation was concentrated on the performance of machine learning based models in aspect of linear-discriminant-analysis (LDA) and non-LDA (logic based). In the group of LDA, na$\"{\i}$ve baysian, k-nearest neighbor, and SVM were used, in the group of non LDA, decision tree, random forest, and boost were used. We described the details of feature vector construction and the selection processes of the model and the parameter including training and validation. We also presented the experimental results of comparison on training/classification error levels for the models employed.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.5
• /
• pp.13-20
• /
• 2016

This study explores the effects of an video-recording English speaking task model on learners. The learning model, a form of mobile learning, was developed to facilitate the learners' output practice applying advantages of a smartphone and Text-to Speech. The survey results shows the positive effects of the speaking task on the domain of pronunciation, speaking, listening, writing in terms of students' confidence, as well as general English ability. The study further examines the possibilities and limitations of the speaking task in assisting Korean learners improve their speaking ability, who do not have sufficient exposure to English input or output practice due to the situational limitations where English is learned as a foreign language.

• Journal of the Korea Institute of Military Science and Technology
• /
• v.25 no.1
• /
• pp.55-64
• /
• 2022

Identifying warships contacted at sea is important to prepare for threats. It is necessary to obtain a basis to identify warships. In this study, we propose a 2-step model that acquires the warship's type and hullnumber with identification information from the warship images. The model classifies the warship's type and detects its hullnumber area by applying object detection, then recognizes hullnumber through text recognition algorithms. Proposed model achieved high performance by using state-of-the-art deep learning algorithms.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2019.07a
• /
• pp.385-386
• /
• 2019

This paper proposes a fusion model based on Long-Short Term Memory networks (LSTM) and CNN deep learning methods, and applied to multi-category news datasets, and achieved good results. Experiments show that the fusion model based on deep learning has greatly improved the precision and accuracy of text sentiment classification.

• Journal of Information Processing Systems
• /
• v.19 no.3
• /
• pp.289-301
• /
• 2023

Due to the development and dissemination of modern technology, anyone can easily communicate using services such as social network service (SNS) through a personal computer (PC) or smartphone. The development of these technologies has caused many beneficial effects. At the same time, bad effects also occurred, one of which was the spam problem. Spam refers to unwanted or rejected information received by unspecified users. The continuous exposure of such information to service users creates inconvenience in the user's use of the service, and if filtering is not performed correctly, the quality of service deteriorates. Recently, spammers are creating more malicious spam by distorting the image of spam text so that optical character recognition (OCR)-based spam filters cannot easily detect it. Fortunately, the level of transformation of image spam circulated on social media is not serious yet. However, in the mail system, spammers (the person who sends spam) showed various modifications to the spam image for neutralizing OCR, and therefore, the same situation can happen with spam images on social media. Spammers have been shown to interfere with OCR reading through geometric transformations such as image distortion, noise addition, and blurring. Various techniques have been studied to filter image spam, but at the same time, methods of interfering with image spam identification using obfuscated images are also continuously developing. In this paper, we propose a deep learning-based spam image detection model to improve the existing OCR-based spam image detection performance and compensate for vulnerabilities. The proposed model extracts text features and image features from the image using four sub-models. First, the OCR-based text model extracts the text-related features, whether the image contains spam words, and the word embedding vector from the input image. Then, the convolution neural network-based image model extracts image obfuscation and image feature vectors from the input image. The extracted feature is determined whether it is a spam image by the final spam image classifier. As a result of evaluating the F1-score of the proposed model, the performance was about 14 points higher than the OCR-based spam image detection performance.

• ETRI Journal
• /
• v.43 no.1
• /
• pp.95-108
• /
• 2021

Application (app) ratings are feedback provided voluntarily by users and serve as important evaluation criteria for apps. However, these ratings can often be biased owing to insufficient or missing votes. Additionally, significant differences have been observed between numeric ratings and user reviews. This study aims to predict the numeric ratings of Google apps using machine learning classifiers. It exploits numeric app ratings provided by users as training data and returns authentic mobile app ratings by analyzing user reviews. An ensemble learning model is proposed for this purpose that considers term frequency/inverse document frequency (TF/IDF) features. Three TF/IDF features, including unigrams, bigrams, and trigrams, were used. The dataset was scraped from the Google Play store, extracting data from 14 different app categories. Biased and unbiased user ratings were discriminated using TextBlob analysis to formulate the ground truth, from which the classifier prediction accuracy was then evaluated. The results demonstrate the high potential for machine learning-based classifiers to predict authentic numeric ratings based on actual user reviews.

• English Language & Literature Teaching
• /
• v.11 no.2
• /
• pp.19-34
• /
• 2005

Hypertext raises question to general assumptions about our conventional conceptions of education. In this essay, three kinds of learning-models are presented by the application of "writing" hypertext literature to the English education of the elementary school. These models, which I call the "scene-centered" system, give knowledge to learners in non-linear, non-sequential structure. The term "scene" is a single concept or idea composed of a single sub-text, which is to be made by the group of students. This system is focused on the collaborative composition of students. Students, by generating sub-texts and connecting texts, perform the educational activities to expand the source text. The "scene-centered" system is, to put it into a Barte's term, a "writerly text." But in order to "write," "reading" should be accompanied. So, this system is a learning model in which writing and reading are carried on simultaneously. In all the process, students play a role of multi-user, with three access rights: read, write, and annotate. So, students making use of hypertext systems will act as reader-authors. And teachers will take the new role in collaborative writing environment. No longer the central authoritarian evaluator, they will become consultants, co-writers, coaches of their students.

• Journal of Information Technology Services
• /
• v.21 no.2
• /
• pp.109-126
• /
• 2022

The number and variety of products and services offered by companies have increased dramatically, providing customers with more choices to meet their needs. As a solution to this information overload problem, the provision of tailored services to individuals has become increasingly important, and the personalized recommender systems have been widely studied and used in both academia and industry. Existing recommender systems face important problems in practical applications. The most important problem is that it cannot clearly explain why it recommends these products. In recent years, some researchers have found that the explanation of recommender systems may be very useful. As a result, users are generally increasing conversion rates, satisfaction, and trust in the recommender system if it is explained why those particular items are recommended. Therefore, this study presents a methodology of providing an explanatory function of a recommender system using a review text left by a user. The basic idea is not to use all of the user's reviews, but to provide them in a summarized form using only reviews left by similar users or neighbors involved in recommending the item as an explanation when providing the recommended item to the user. To achieve this research goal, this study aims to provide a product recommendation list using user-based collaborative filtering techniques, combine reviews left by neighboring users with each product to build a model that combines text summary methods among deep learning-based natural language processing methods. Using the IMDb movie database, text reviews of all target user neighbors' movies are collected and summarized to present descriptions of recommended movies. There are several text summary methods, but this study aims to evaluate whether the review summary is well performed by training the Sequence-to-sequence+attention model, which is a representative generation summary method, and the BertSum model, which is an extraction summary model.

• Journal of Information Processing Systems
• /
• v.17 no.4
• /
• pp.818-833
• /
• 2021

To address the problems of low precision rate, insufficient feature extraction, and poor contextual ability in existing text sentiment analysis methods, a mixed model account of a CNN-BiLSTM-TE (convolutional neural network, bidirectional long short-term memory, and topic extraction) model was proposed. First, Chinese text data was converted into vectors through the method of transfer learning by Word2Vec. Second, local features were extracted by the CNN model. Then, contextual information was extracted by the BiLSTM neural network and the emotional tendency was obtained using softmax. Finally, topics were extracted by the term frequency-inverse document frequency and K-means. Compared with the CNN, BiLSTM, and gate recurrent unit (GRU) models, the CNN-BiLSTM-TE model's F1-score was higher than other models by 0.0147, 0.006, and 0.0052, respectively. Then compared with CNN-LSTM, LSTM-CNN, and BiLSTM-CNN models, the F1-score was higher by 0.0071, 0.0038, and 0.0049, respectively. Experimental results showed that the CNN-BiLSTM-TE model can effectively improve various indicators in application. Lastly, performed scalability verification through a takeaway dataset, which has great value in practical applications.