• Journal of Intelligence and Information Systems
• /
• v.28 no.4
• /
• pp.135-155
• /
• 2022

This study analyzes patent trends by deriving sub-technical fields of Organic Light-Emitting Diodes (OLEDs) industry, and analyzing technology value, originality, and diversity for each sub-technical field. To collect patent data, a set of international patent classification(IPC) codes related to OLED technology was defined, and OLED-related patents applied from 2005 to 2017 were collected using a set of IPC codes. Then, a large number of collected patent documents were classified into 12 major technologies using the Latent Dirichlet Allocation(LDA) topic model and trends for each technology were investigated. Patents related to touch sensor, module, image processing, and circuit driving showed an increasing trend, but virtual reality and user interface recently decreased, and thin film transistor, fingerprint recognition, and optical film showed a continuous trend. To compare the technological value, the number of forward citations, originality, and diversity of patents included in each technology group were investigated. From the results, image processing, user interface(UI) and user experience(UX), module, and adhesive technology with high number of forward citations, originality and diversity showed relatively high technological value. The results provide useful information in the process of establishing a company's technology strategy.

• Design Convergence Study
• /
• v.15 no.2
• /
• pp.347-368
• /
• 2016

Recently, researches for the word of mouth(WOM) imply that consumers use WOM informations of products in their purchase process. This study suggests methods using opinion mining and visualization to understand consumers' opinion of each goods and each markets. For this study we conduct research that includes developing domain ontology based on reviews confined to "movie" category because people who want to have watching movie refer other's movie reviews recently, and it is analyzed by opinion mining and visualization. It has differences comparing other researches as conducting attribution classification of evaluation factors and comprising verbal dictionary about evaluation factors when we conduct ontology process for analyzing. We want to prove through the result if research method will be valid. Results derived from this study can be largely divided into three. First, This research explains methods of developing domain ontology using keyword extraction and topic modeling. Second, We visualize reviews of each movie to understand overall audiences' opinion about specific movies. Third, We find clusters that consist of products which evaluated similar assessments in accordance with the evaluation results for the product. Case study of this research largely shows three clusters containing 130 movies that are used according to audiences'opinion.

• Journal of Intelligence and Information Systems
• /
• v.29 no.4
• /
• pp.309-323
• /
• 2023

As the role of online reviews has become increasingly crucial, numerous studies have been conducted to utilize helpful reviews. Helpful reviews, perceived by customers, have been verified in various research studies to be influenced by factors such as ratings, review length, review content, and so on. The determination of a review's helpfulness is generally based on the number of 'helpful' votes from consumers, with more 'helpful' votes considered to have a more significant impact on consumers' purchasing decisions. However, recently written reviews that have not been exposed to many customers may have relatively few 'helpful' votes and may lack 'helpful' votes altogether due to a lack of participation. Therefore, rather than relying on the number of 'helpful' votes to assess the helpfulness of reviews, we aim to classify them based on review content. In addition, the text of the review emerges as the most influential factor in review helpfulness. This study employs text mining techniques, including topic modeling and sentiment analysis, to analyze the diverse impacts of content and emotions embedded in the review text. In this study, we propose a review helpfulness prediction model based on review content, utilizing movie reviews from IMDb, a global movie information site. We construct a review helpfulness prediction model by using an explainable Graph Neural Network (GNN), while addressing the interpretability limitations of the machine learning model. The explainable graph neural network is expected to provide more reliable information about helpful or non-helpful reviews as it can identify connections between reviews.

• Journal of the Korean Society for Library and Information Science
• /
• v.58 no.2
• /
• pp.363-384
• /
• 2024

This study was conducted to examine the user's experiences with the public library in Daejeon using big data analysis, focusing on the text mining technique. To know this, first, the overall evaluation and perception of users about the public library in Daejeon were explored by collecting data on social media. Second, through analysis using online news articles, the pending issues that are being discussed socially were identified. As a result of the analysis, the proportion of users with children was first high. Next, it was found that topics through LDA analysis appeared in four categories: 'cultural event/program', 'data use', 'physical environment and facilities', and 'library service'. Finally, it was confirmed that keywords for the additional construction of libraries and complex cultural spaces and the establishment of a library cooperation system appeared at the core in the news article data. Based on this, it was proposed to build a library in consideration of regional balance and to create a social parenting community network through business agreements with childcare and childcare institutions. This will contribute to identifying the policy and social trends of public libraries in Daejeon and implementing data-based public library operations that reflect local community demands.

• 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.