• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.191-206
• /
• 2022

Recently, it is a de-facto approach to utilize a pre-trained language model(PLM) to achieve the state-of-the-art performance for various natural language tasks(called downstream tasks) such as sentiment analysis and question answering. However, similar to any other machine learning method, PLM tends to depend on the data distribution seen during the training phase and shows worse performance on the unseen (Out-of-Distribution) domain. Due to the aforementioned reason, there have been many efforts to develop domain-specified PLM for various fields such as medical and legal industries. In this paper, we discuss the training of a finance domain-specified PLM for the Korean language and its applications. Our finance domain-specified PLM, KB-BERT, is trained on a carefully curated financial corpus that includes domain-specific documents such as financial reports. We provide extensive performance evaluation results on three natural language tasks, topic classification, sentiment analysis, and question answering. Compared to the state-of-the-art Korean PLM models such as KoELECTRA and KLUE-RoBERTa, KB-BERT shows comparable performance on general datasets based on common corpora like Wikipedia and news articles. Moreover, KB-BERT outperforms compared models on finance domain datasets that require finance-specific knowledge to solve given problems.

• Journal of Nutrition and Health
• /
• v.55 no.1
• /
• pp.21-35
• /
• 2022

The recommended meal composition allows the general people to organize meals using the number of intakes of foods from each of six food groups (grains, meat·fish·eggs·beans, vegetables, fruits, milk·dairy products and oils·sugars) to meet Dietary Reference Intakes for Koreans (KDRIs) without calculating complex nutritional values. Through an integrated analysis of data from the 6th to 7th Korean National Health and Nutrition Examination Surveys (2013-2018), representative foods for each food group were selected, and the amounts of representative foods per person were derived based on energy. Based on the EER by age and gender from the KDRIs, a total of 12 kinds of diets were suggested by differentiating meal compositions by age (aged 1-2, 3-5, 6-11, 12-18, 19-64, 65-74 and ≥ 75 years) and gender. The 2020 Food Balance Wheel included the 6th food group of oils and sugars to raise public awareness and avoid confusion in the practical utilization of the model by industries or individuals in reducing the consistent increasing intakes of oils and sugars. To promote the everyday use of the Food Balance Wheel and recommended meal compositions among the general public, the poster of the Food Balance Wheel was created in five languages (Korean, English, Japanese, Vietnamese and Chinese) along with card news. A survey was conducted to provide a basis for categorizing nutritional problems by life cycles and developing customized web-based messages to the public. Based on survey results two types of card news were produced for the general public and youth. Additionally, the educational program was developed through a series of processes, such as prioritization of educational topics, setting educational goals for each stage, creation of a detailed educational system chart and teaching-learning plans for the development of educational materials and media.

• Journal of Intelligence and Information Systems
• /
• v.25 no.3
• /
• pp.161-177
• /
• 2019

In this paper, we study the performance improvement of the answer extraction in Question-Answering system by using sentence dependency parsing result. The Question-Answering (QA) system consists of query analysis, which is a method of analyzing the user's query, and answer extraction, which is a method to extract appropriate answers in the document. And various studies have been conducted on two methods. In order to improve the performance of answer extraction, it is necessary to accurately reflect the grammatical information of sentences. In Korean, because word order structure is free and omission of sentence components is frequent, dependency parsing is a good way to analyze Korean syntax. Therefore, in this study, we improved the performance of the answer extraction by adding the features generated by dependency parsing analysis to the inputs of the answer extraction model (Bidirectional LSTM-CRF). The process of generating the dependency graph embedding consists of the steps of generating the dependency graph from the dependency parsing result and learning the embedding of the graph. In this study, we compared the performance of the answer extraction model when inputting basic word features generated without the dependency parsing and the performance of the model when inputting the addition of the Eojeol tag feature and dependency graph embedding feature. Since dependency parsing is performed on a basic unit of an Eojeol, which is a component of sentences separated by a space, the tag information of the Eojeol can be obtained as a result of the dependency parsing. The Eojeol tag feature means the tag information of the Eojeol. The process of generating the dependency graph embedding consists of the steps of generating the dependency graph from the dependency parsing result and learning the embedding of the graph. From the dependency parsing result, a graph is generated from the Eojeol to the node, the dependency between the Eojeol to the edge, and the Eojeol tag to the node label. In this process, an undirected graph is generated or a directed graph is generated according to whether or not the dependency relation direction is considered. To obtain the embedding of the graph, we used Graph2Vec, which is a method of finding the embedding of the graph by the subgraphs constituting a graph. We can specify the maximum path length between nodes in the process of finding subgraphs of a graph. If the maximum path length between nodes is 1, graph embedding is generated only by direct dependency between Eojeol, and graph embedding is generated including indirect dependencies as the maximum path length between nodes becomes larger. In the experiment, the maximum path length between nodes is adjusted differently from 1 to 3 depending on whether direction of dependency is considered or not, and the performance of answer extraction is measured. Experimental results show that both Eojeol tag feature and dependency graph embedding feature improve the performance of answer extraction. In particular, considering the direction of the dependency relation and extracting the dependency graph generated with the maximum path length of 1 in the subgraph extraction process in Graph2Vec as the input of the model, the highest answer extraction performance was shown. As a result of these experiments, we concluded that it is better to take into account the direction of dependence and to consider only the direct connection rather than the indirect dependence between the words. The significance of this study is as follows. First, we improved the performance of answer extraction by adding features using dependency parsing results, taking into account the characteristics of Korean, which is free of word order structure and omission of sentence components. Second, we generated feature of dependency parsing result by learning - based graph embedding method without defining the pattern of dependency between Eojeol. Future research directions are as follows. In this study, the features generated as a result of the dependency parsing are applied only to the answer extraction model in order to grasp the meaning. However, in the future, if the performance is confirmed by applying the features to various natural language processing models such as sentiment analysis or name entity recognition, the validity of the features can be verified more accurately.

• Journal of Intelligence and Information Systems
• /
• v.21 no.2
• /
• pp.49-67
• /
• 2015

Recently, emerging the notion of big data and social media has led us to enter data's big bang. Social networking services are widely used by people around the world, and they have become a part of major communication tools for all ages. Over the last decade, as online social networking sites become increasingly popular, companies tend to focus on advanced social media analysis for their marketing strategies. In addition to social media analysis, companies are mainly concerned about propagating of negative opinions on social networking sites such as Facebook and Twitter, as well as e-commerce sites. The effect of online word of mouth (WOM) such as product rating, product review, and product recommendations is very influential, and negative opinions have significant impact on product sales. This trend has increased researchers' attention to a natural language processing, such as a sentiment analysis. A sentiment analysis, also refers to as an opinion mining, is a process of identifying the polarity of subjective information and has been applied to various research and practical fields. However, there are obstacles lies when Korean language (Hangul) is used in a natural language processing because it is an agglutinative language with rich morphology pose problems. Therefore, there is a lack of Korean natural language processing resources such as a sentiment lexicon, and this has resulted in significant limitations for researchers and practitioners who are considering sentiment analysis. Our study builds a Korean sentiment lexicon with collective intelligence, and provides API (Application Programming Interface) service to open and share a sentiment lexicon data with the public (www.openhangul.com). For the pre-processing, we have created a Korean lexicon database with over 517,178 words and classified them into sentiment and non-sentiment words. In order to classify them, we first identified stop words which often quite likely to play a negative role in sentiment analysis and excluded them from our sentiment scoring. In general, sentiment words are nouns, adjectives, verbs, adverbs as they have sentimental expressions such as positive, neutral, and negative. On the other hands, non-sentiment words are interjection, determiner, numeral, postposition, etc. as they generally have no sentimental expressions. To build a reliable sentiment lexicon, we have adopted a concept of collective intelligence as a model for crowdsourcing. In addition, a concept of folksonomy has been implemented in the process of taxonomy to help collective intelligence. In order to make up for an inherent weakness of folksonomy, we have adopted a majority rule by building a voting system. Participants, as voters were offered three voting options to choose from positivity, negativity, and neutrality, and the voting have been conducted on one of the largest social networking sites for college students in Korea. More than 35,000 votes have been made by college students in Korea, and we keep this voting system open by maintaining the project as a perpetual study. Besides, any change in the sentiment score of words can be an important observation because it enables us to keep track of temporal changes in Korean language as a natural language. Lastly, our study offers a RESTful, JSON based API service through a web platform to make easier support for users such as researchers, companies, and developers. Finally, our study makes important contributions to both research and practice. In terms of research, our Korean sentiment lexicon plays an important role as a resource for Korean natural language processing. In terms of practice, practitioners such as managers and marketers can implement sentiment analysis effectively by using Korean sentiment lexicon we built. Moreover, our study sheds new light on the value of folksonomy by combining collective intelligence, and we also expect to give a new direction and a new start to the development of Korean natural language processing.

• Journal of Intelligence and Information Systems
• /
• v.21 no.2
• /
• pp.69-92
• /
• 2015

The explosion of social media data has led to apply text-mining techniques to analyze big social media data in a more rigorous manner. Even if social media text analysis algorithms were improved, previous approaches to social media text analysis have some limitations. In the field of sentiment analysis of social media written in Korean, there are two typical approaches. One is the linguistic approach using machine learning, which is the most common approach. Some studies have been conducted by adding grammatical factors to feature sets for training classification model. The other approach adopts the semantic analysis method to sentiment analysis, but this approach is mainly applied to English texts. To overcome these limitations, this study applies the Word2Vec algorithm which is an extension of the neural network algorithms to deal with more extensive semantic features that were underestimated in existing sentiment analysis. The result from adopting the Word2Vec algorithm is compared to the result from co-occurrence analysis to identify the difference between two approaches. The results show that the distribution related word extracted by Word2Vec algorithm in that the words represent some emotion about the keyword used are three times more than extracted by co-occurrence analysis. The reason of the difference between two results comes from Word2Vec's semantic features vectorization. Therefore, it is possible to say that Word2Vec algorithm is able to catch the hidden related words which have not been found in traditional analysis. In addition, Part Of Speech (POS) tagging for Korean is used to detect adjective as "emotional word" in Korean. In addition, the emotion words extracted from the text are converted into word vector by the Word2Vec algorithm to find related words. Among these related words, noun words are selected because each word of them would have causal relationship with "emotional word" in the sentence. The process of extracting these trigger factor of emotional word is named "Emotion Trigger" in this study. As a case study, the datasets used in the study are collected by searching using three keywords: professor, prosecutor, and doctor in that these keywords contain rich public emotion and opinion. Advanced data collecting was conducted to select secondary keywords for data gathering. The secondary keywords for each keyword used to gather the data to be used in actual analysis are followed: Professor (sexual assault, misappropriation of research money, recruitment irregularities, polifessor), Doctor (Shin hae-chul sky hospital, drinking and plastic surgery, rebate) Prosecutor (lewd behavior, sponsor). The size of the text data is about to 100,000(Professor: 25720, Doctor: 35110, Prosecutor: 43225) and the data are gathered from news, blog, and twitter to reflect various level of public emotion into text data analysis. As a visualization method, Gephi (http://gephi.github.io) was used and every program used in text processing and analysis are java coding. The contributions of this study are as follows: First, different approaches for sentiment analysis are integrated to overcome the limitations of existing approaches. Secondly, finding Emotion Trigger can detect the hidden connections to public emotion which existing method cannot detect. Finally, the approach used in this study could be generalized regardless of types of text data. The limitation of this study is that it is hard to say the word extracted by Emotion Trigger processing has significantly causal relationship with emotional word in a sentence. The future study will be conducted to clarify the causal relationship between emotional words and the words extracted by Emotion Trigger by comparing with the relationships manually tagged. Furthermore, the text data used in Emotion Trigger are twitter, so the data have a number of distinct features which we did not deal with in this study. These features will be considered in further study.