• Journal of Korea Society of Industrial Information Systems
• /
• v.28 no.5
• /
• pp.15-30
• /
• 2023

Recently, generative models based on the Transformer architecture, such as ChatGPT, have been gaining significant attention. The Transformer architecture has been applied to various neural network models, including Google's BERT(Bidirectional Encoder Representations from Transformers) sentence generation model. In this paper, a method is proposed to create a text binary classification model for determining whether a comment on Korean movie review is positive or negative. To accomplish this, a pre-trained multilingual BERT sentence generation model is fine-tuned and transfer learned using a new Korean training dataset. To achieve this, a pre-trained BERT-Base model for multilingual sentence generation with 104 languages, 12 layers, 768 hidden, 12 attention heads, and 110M parameters is used. To change the pre-trained BERT-Base model into a text classification model, the input and output layers were fine-tuned, resulting in the creation of a new model with 178 million parameters. Using the fine-tuned model, with a maximum word count of 128, a batch size of 16, and 5 epochs, transfer learning is conducted with 10,000 training data and 5,000 testing data. A text sentiment binary classification model for Korean movie review with an accuracy of 0.9582, a loss of 0.1177, and an F1 score of 0.81 has been created. As a result of performing transfer learning with a dataset five times larger, a model with an accuracy of 0.9562, a loss of 0.1202, and an F1 score of 0.86 has been generated.

• Cartoon and Animation Studies
• /
• s.34
• /
• pp.345-366
• /
• 2014

From the beginning of Digital text book in 2007, there are many kinds of digital text book such as English, Science etc at Public education. Above many problems at the beginning like just using paper text book's scan data as digital text book, now use special contents and design for only digital text book. But only for digital text book not for other. There is gap between advanced country of education and us. This is research based on LG europe design center in London, UK target is UK, Sweden by heuristic analysis, question investigation to get Target's UX with digital education media. Advancement of digital and interest of education bring the world development of digital education device. UK, where is education advanced nation, is using lot's of digital education device which is interactive board, digital desk etc. Result of Analysis of Digital Education Design trend by Target of Education Needs apply rough Design by LG europe design center. We can get more sophisticated needs and UX result by target then Korea that can use for our future Digital education design plan. Also help to reduce gap between advanced country and Korea.

• Journal of KIISE:Software and Applications
• /
• v.29 no.4
• /
• pp.270-277
• /
• 2002

Text filtering is a task of deciding whether a document has relevance to a specified topic. As Internet and Web becomes wide-spread and the number of documents delivered by e-mail explosively grows the importance of text filtering increases as well. The aim of this paper is to improve the accuracy of text filtering systems by using machine learning techniques. We apply AdaBoost algorithms to the filtering task. An AdaBoost algorithm generates and combines a series of simple hypotheses. Each of the hypotheses decides the relevance of a document to a topic on the basis of whether or not the document includes a certain word. We begin with an existing AdaBoost algorithm which uses weak hypotheses with their output of 1 or -1. Then we extend the algorithm to use weak hypotheses with real-valued outputs which was proposed recently to improve error reduction rates and final filtering performance. Next, we attempt to achieve further improvement in the AdaBoost's performance by first setting weights randomly according to the continuous Poisson distribution, executing AdaBoost, repeating these steps several times, and then combining all the hypotheses learned. This has the effect of mitigating the ovefitting problem which may occur when learning from a small number of data. Experiments have been performed on the real document collections used in TREC-8, a well-established text retrieval contest. This dataset includes Financial Times articles from 1992 to 1994. The experimental results show that AdaBoost with real-valued hypotheses outperforms AdaBoost with binary-valued hypotheses, and that AdaBoost iterated with random weights further improves filtering accuracy. Comparison results of all the participants of the TREC-8 filtering task are also provided.

• The KIPS Transactions:PartB
• /
• v.15B no.3
• /
• pp.179-188
• /
• 2008

In this paper, knowledge based text to facial sequence image system for interaction of lecturer and learner in cyber universities is studied. The system is defined by the synthesis of facial sequence image which is synchronized the lip according to the text information based on grammatical characteristic of hangul. For the implementation of the system, the transformation method that the text information is transformed into the phoneme code, the deformation rules of mouse shape which can be changed according to the code of phonemes, and the synthesis method of facial sequence image by using deformation rules of mouse shape are proposed. In the proposed method, all syllables of hangul are represented 10 principal mouse shape and 78 compound mouse shape according to the pronunciation characteristics of the basic consonants and vowels, and the characteristics of the articulation rules, respectively. To synthesize the real time facial sequence image able to realize the PC, the 88 mouth shape stored data base are used without the synthesis of mouse shape in each frame. To verify the validity of the proposed method the various synthesis of facial sequence image transformed from the text information is accomplished, and the system that can be applied the PC is implemented using the proposed method.

• Journal of Information Technology Services
• /
• v.14 no.1
• /
• pp.237-249
• /
• 2015

The volume of unstructured text data generated by various social media has been increasing rapidly; therefore, use of text mining to support decision making has also been increasing. Especially, issue Clustering-determining a new relation with various issues through clustering-has gained attention from many researchers. However, traditional issue clustering methods can only be performed based on the co-occurrence frequency of issue keywords in many documents. Therefore, an association between issues that have a low co-occurrence frequency cannot be discovered using traditional issue clustering methods, even if those issues are strongly related in other perspectives. Therefore, issue clustering that fits each of criteria needs to be performed by the perspective of analysis and the purpose of use. In this study, a multi-dimensional issue clustering is proposed to overcome the limitation of traditional issue clustering. We assert, specifically in this study, that issue clustering should be performed for a particular purpose. We analyze the results of applying our methodology to two specific perspectives on issue clustering, (i) consumers' interests, and (ii) related R&D terms.

• Journal of Information Technology Services
• /
• v.14 no.2
• /
• pp.159-176
• /
• 2015

As a public medium and one of the big data sources that is accumulated informally and real time, online news comments or replies are considered a significant resource to understand mentalities of article readers. The comments are also being regarded as an important medium of WOM (Word of Mouse) about products, services or the enterprises. If the diffusing effect of the comments is referred to as the degrees of agreement and disagreement from an angle of WOM, figuring out which characteristics of the comments would influence the agreements or the disagreements to the comments in very early stage would be very worthwhile to establish a comment-based eWOM (electronic WOM) strategy. However, investigating the effects of the characteristics of the comments on eWOM effect has been rarely studied. According to this angle, this study aims to conduct an empirical analysis which understands the characteristics of comments that affect the numbers of agreement and disagreement, as eWOM performance, to particular news articles which address a specific product, service or enterprise per se. While extant literature has focused on the quantitative attributes of the comments which are collected by manually, this paper used text mining techniques to acquire the qualitative attributes of the comments in an automatic and cost effective manner.

• Journal of Information Technology Services
• /
• v.17 no.1
• /
• pp.79-89
• /
• 2018

Sentiment analysis is a technique of text mining that extracts feelings of the person who wrote the sentence like movie review. The preliminary researches of sentiment analysis identify sentiments by using the dictionary which contains negative and positive words collected in advance. As researches on deep learning are actively carried out, sentiment analysis using deep learning model with morpheme or word unit has been done. However, this model has disadvantages in that the word dictionary varies according to the domain and the number of morphemes or words gets relatively larger than that of phonemes. Therefore, the size of the dictionary becomes large and the complexity of the model increases accordingly. We construct a sentiment analysis model using recurrent neural network by dividing input data into phoneme-level which is smaller than morpheme-level. To verify the performance, we use 30,000 movie reviews from the Korean biggest portal, Naver. Morpheme-level sentiment analysis model is also implemented and compared. As a result, the phoneme-level sentiment analysis model is superior to that of the morpheme-level, and in particular, the phoneme-level model using LSTM performs better than that of using GRU model. It is expected that Korean text processing based on a phoneme-level model can be applied to various text mining and language models.

• Journal of the Korea Computer Industry Society
• /
• v.2 no.12
• /
• pp.1601-1608
• /
• 2001

Digital library is the most important database system that needs information retrieval engine for natural language documents and multimedia data. This paper describes the experimental results of information retrieval engine and browser based on natural language processing. It includes lexical analysis, syntax processing, stemming, and keyword indexing for the natural language text. With the experimental database ‘Earth and Space Science’ that has lots of images and titles and their descriptive text in natural language, text-based search engine was tested. Combined with content-based image search engine, it is expected to be a multimedia information retrieval system in digital library

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2003.10a
• /
• pp.86-94
• /
• 2003

Electronically available biological literature has been accumulated exponentially in the course of time. So, researches on automatically acquiring knowledge from these tremendous data by text mining technology become more and more prosperous. However, most of the previous researches are technology oriented and are not well focused in practical extraction target, hence result in low performance and inconvenience for the bio-researchers to actually use. In this paper, we propose a more biology oriented target domain specific text mining system, that is, POSTECH bio-text mining system (POSBIOTM), for signal transduction pathway extraction, especially for G protein-coupled receptor (GPCR) pathway. To reflect more domain knowledge, we specify the concrete target for pathway extraction and define the minimal pathway domain ontology. Under this conceptual model, POSBIOTM extracts interactions and entities of pathways from the full biological articles using a machine learning oriented extraction method and visualizes the pathways using JDesigner module provided in the system biology workbench (SBW) [14]

• Knowledge Management Research
• /
• v.19 no.2
• /
• pp.91-108
• /
• 2018

This study mainly focused on measuring the impact of comments for a particular song on the number of streamings and downloads. We modeled multiple regression equations to perform this analysis. We chose digital music market for the object of analysis because of its inherent characteristics, such as experience goods, high bandwagon effect, and so on. We carefully utilized text mining technique in accordance with the algorithm of Naïve Bayes classifier to distinguish whether a comment for a piece of music be regarded as positive or negative. In addition, we used 'size of agency' and 'existence of hit song' as moderating variables. The reason for usage of those variables is that those are assumed to affect users' decision for selecting particular song especially when downloading or streaming via music sites. We found empirical evidences that positive comments for a particular song increase the number of both downloads and streamings. However, positive comments may decrease the number of downloads when the size of agency of the artist is big. As a result, we were able to say that a positive comment for a particular song functioned as 'word-of-mouth' effect, inducing other users' behavioral response. We also found that other features of an artist such as size of the agency that the artist belongs to functioned as an external factor along with feature of the song itself.