• Asia-pacific Journal of Multimedia Services Convergent with Art, Humanities, and Sociology
• /
• v.8 no.3
• /
• pp.531-541
• /
• 2018

This study researched text mining and text network analysis to analyze the images of Nonghyup Insurance for consumers. With the recent development of social media, many texts are being produced and reproduced, and texts of social media provide important information to companies. Text mining and text network analysis are used in many studies to identify image of company and product. As a result of the text analysis, the positive image of the Nonghyup Insurance is safety and stability. Negative images of the Nonghyup Insurance is concern and anxiety. As a result of the textual network analysis, Centered mage of Nonghyup Insurance is safety and concern. This paper allows researchers to extract several lessons learned that are important for the text mining and text network analysis.

• The Journal of the Korea Contents Association
• /
• v.17 no.10
• /
• pp.492-502
• /
• 2017

This study, by analyzing two recent television dramas, attempts to identify the ways how popular cultural texts deal with supernatural beings and to discuss political meanings of the ways in the context of neoliberal Korea. The results are: (1) The narratives make a clear line between the supernatural and the ordinary. (2) The supernatural is effectively used in the narratives to extend the boundary of conflict structure towards structural social problems that the society is now facing. (3) When the text resolving the conflicts, the supernatural is also critical, which makes the whle narrative in line with fantasy rather than reality. These results suggest that the conclusion of the previous studies, arguing the use of the supernatural by popular cultural texts tends to function as a form of resistance against neoliberal discourse structure, needs to be negotiated. It is also reaffirmed that we need to explain political meanings of popular cultural texts dealing with supernatural beings, with its double-sided and ambivalent effects.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2009.01a
• /
• pp.773-777
• /
• 2009

In recent years, the field of synthesizing voice has been developed rapidly, and the technologies such as reading aloud an email or sound guidance of a car navigation system are used in various scenes of our life. The sound quality is monotonous like reading news. It is preferable for a text such as a novel to be read by the voice that expresses emotions wealthily. Therefore, we have been trying to develop a system reading aloud novels automatically that are expressed clear emotions comparatively such as juvenile literature. At first it is necessary to identify emotions expressed in a sentence in texts in order to make a computer read texts with an emotionally expressive voice. A method on the basis of the meaning interpretation that utilized artificial intelligence technology for a method to specify emotions of texts is thought, but it is very difficult with the current technology. Therefore, we propose a method to determine only emotion every sentence in a novel by a simpler way. This method determines the emotion of a sentence according to an emotion that words such as a verb in a Japanese verb sentence, and an adjective and an adverb in a adjective sentence, have. The emotional characteristics that these words have are prepared beforehand as a emotional words dictionary by us. The emotions used here are seven types: "joy," "sorrow," "anger," "surprise," "terror," "aversion" or "neutral."

• Journal of Digital Convergence
• /
• v.18 no.9
• /
• pp.17-27
• /
• 2020

News articles and blog posts have influence on social agenda setting and this study applied text mining on the subject of solar PV in rural area appeared in those media. Texts are gained from online news articles and blog posts with rural solar PV as a keyword by web scrapping, and these are analysed by sentiment analysis and topic modeling technique. Sentiment analysis shows that the proportion of negative texts are significantly lower in blog posts compared to news articles. Result of topic modeling shows that topics related to government policy have the largest loading in positive articles whereas various topics are relatively evenly distributed in negative articles. For blog posts, topics related to rural area installation and environmental damage are have the largest loading in positive and negative texts, respectively. This research reveals issues related to rural solar PV by combining sentiment analysis and topic modeling that were separately applied in previous studies.

• Journal of Korean Classical Literature and Education
• /
• no.39
• /
• pp.5-39
• /
• 2018

This paper aims to identify the characteristics of literary therapy in relation to literature education. It also intends to clarify its distinctiveness. Literary therapy is not to teach literature. It does not deliver knowledge on agreed analyses, backgrounds and the nature of genres. Literary therapy encourages participants to fully appreciate one's thought and emotions and express them. The end goal is self-knowledge rather than the understanding of texts. Literary therapy focuses on self-knowledge through literatures as opposed to literature education which aims to encourage understandings of literature texts. In literary therapy, literature is media for personal growth facilitating self-expansion. Literature works enable participants to view oneself objectively by the means of one's responses to literature works. Literary therapy has more permissive viewpoints on recipients' response to literature texts than literature education. In addition, the subject of literary therapy is more unique and individualistic.

• International Journal of Advanced Culture Technology
• /
• v.4 no.3
• /
• pp.62-68
• /
• 2016

This study explores evaluative linguistic expressions in news reporting about the 2016 general election outcome in Korean newspapers. In particular, we have examined the evaluative linguistic expressions quoted from the three Western news media -New York Times, Washington Post, and BBC, both quantitatively and qualitatively in Korean news stories in order to know how journalists frame the news stories to persuade news consumers to accept their ideologies. This is based on the assumption that quotation can be a tool in conveying ideologies to news consumers (van Dijk, 1988, Jullian, 2011). To achieve this purpose, we selected ten Korean newspapers which included quotations from the news stories of the three Western media and then analyzed the quoted expressions quantitatively and qualitatively. For a qualitative analysis, evaluative linguistic expressions were analyzed to examine the journalistic stances of the Western news stories, following Martin's (2003) appraisal theory. For a quantitative analysis, a word frequency analysis was conducted to figure out the ratio of quoted words to the whole news texts in Korean newspapers. As a result, it was found that the news stories of BBC and Washington Post were more frequently quoted than that of New York Times when journalists conveyed neutral or positive attitude to the election outcome, thus confirming that evaluative linguistic expressions were functionally employed to convey journalists' ideologies or stances to news readers.

• The Journal of the Korea Contents Association
• /
• v.19 no.12
• /
• pp.249-258
• /
• 2019

Changes in media technolgy affect the competitive status of broadcasting networks as news media. The competitive media environment has pushed broadcasting network news programs to find new ways for leveling their qualitative performance up and rating. This study focuses on the empirical relationship between the two key value, news quality in terms of fairness and in-depthness and news ratings. This study is based on the analysis of broadcasting network news texts and individual news item raitngs. Empirical relationship between news quality factors and ratings was proved positive. But the relationship between the length of news item and rating was proved negative.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2013.10a
• /
• pp.591-594
• /
• 2013

This study examines how OTT service changes the television viewing experience, what are the underlying dimensions that difference of viewing behavior and experiences between OTT service and old television media. The result showed the characteristic of viewing experiences through OTT service that social viewing, channel zapping by sharing information, multi viewing, de-contextualized viewing space. Specifically, television as a family media turned into a personalized media, social viewing generalized so called 'lean forward' with each other using texts and images at the same time. In particular, information on the content and appraisal shared with each other on SNS in real time.

• Genomics & Informatics
• /
• v.19 no.3
• /
• pp.21.1-21.5
• /
• 2021

The use of social media data, like Twitter, for biomedical research has been gradually increasing over the years. With the coronavirus disease 2019 (COVID-19) pandemic, researchers have turned to more non-traditional sources of clinical data to characterize the disease in near-real time, study the societal implications of interventions, as well as the sequelae that recovered COVID-19 cases present. However, manually curated social media datasets are difficult to come by due to the expensive costs of manual annotation and the efforts needed to identify the correct texts. When datasets are available, they are usually very small and their annotations don't generalize well over time or to larger sets of documents. As part of the 2021 Biomedical Linked Annotation Hackathon, we release our dataset of over 120 million automatically annotated tweets for biomedical research purposes. Incorporating best-practices, we identify tweets with potentially high clinical relevance. We evaluated our work by comparing several SpaCy-based annotation frameworks against a manually annotated gold-standard dataset. Selecting the best method to use for automatic annotation, we then annotated 120 million tweets and released them publicly for future downstream usage within the biomedical domain.

• Journal of Korea Society of Industrial Information Systems
• /
• v.29 no.3
• /
• pp.1-12
• /
• 2024

In this paper, we fine-tuned KoAlpaca, a large-scale Korean language model, and implemented a blog text generation system utilizing it. Blogs on social media platforms are widely used as a marketing tool for businesses. We constructed training data of positive reviews through emotion analysis and refinement of collected sponsored blog texts and applied QLoRA for the lightweight training of KoAlpaca. QLoRA is a fine-tuning approach that significantly reduces the memory usage required for training, with experiments in an environment with a parameter size of 12.8B showing up to a 58.8% decrease in memory usage compared to LoRA. To evaluate the generative performance of the fine-tuned model, texts generated from 100 inputs not included in the training data produced on average more than twice the number of words compared to the pre-trained model, with texts of positive sentiment also appearing more than twice as often. In a survey conducted for qualitative evaluation of generative performance, responses indicated that the fine-tuned model's generated outputs were more relevant to the given topics on average 77.5% of the time. This demonstrates that the positive review generation language model for sponsored content in this paper can enhance the efficiency of time management for content creation and ensure consistent marketing effects. However, to reduce the generation of content that deviates from the category of positive reviews due to elements of the pre-trained model, we plan to proceed with fine-tuning using the augmentation of training data.