19 research outputs found
Development of Intelligent Information System for Digital Cultural Contents
No full textA significant amount of digital cultural contents is shared online, but learners do not know where subject matter content is or how to find it. Therefore, there is a need for a service to improve educational quality by effectively providing relevant information in response to searches for content that is useful to learners. This study developed and tested the usability and utility of an intelligent information system that effectively searches and visualizes digital cultural contents. The system collects data on digital cultural contents, automatically classifies them, and creates content triple data to automatically display the results with a 3D timeline, knowledge network map, and keyword relation network map through content search, triple search, and keyword search. We also conducted a survey and in-depth interviews to verify users’ satisfaction with respect to the use and utility of the system. For the experiment, we developed survey questions to measure user satisfaction and conducted in-depth interviews regarding the system’s utility with a total of 65 subjects. The results show that the response for satisfaction with regard to the use and utility was generally “satisfied”. In addition, the system stability was evaluated as “high”
A Study of Keywords Based on the Word Frequency Effect Theory in Video Lectures of Software Engineering Education for Detecting Mind
No full textComparative Analysis of Learning Effect on Lexical Recognition in the e-Learning and s-Learning
No full textEnhancing Code Similarity with Augmented Data Filtering and Ensemble Strategies
Full text linkAlthough COVID-19 has severely affected the global economy, information technology (IT) employees managed to perform most of their work from home. Telecommuting and remote work have promoted a demand for IT services in various market sectors, including retail, entertainment, education, and healthcare. Consequently, computer and information experts are also in demand. However, producing IT, experts is difficult during a pandemic owing to limitations, such as the reduced enrollment of international students. Therefore, researching increasing software productivity is essential; this study proposes a code similarity determination model that utilizes augmented data filtering and ensemble strategies. This algorithm is the first automated development system for increasing software productivity that addresses the current situation—a worldwide shortage of software dramatically improves performance in various downstream natural language processing tasks (NLP). Unlike general-purpose pre-trained language models (PLMs), CodeBERT and GraphCodeBERT are PLMs that have learned both natural and programming languages. Hence, they are suitable as code similarity determination models. The data filtering process consists of three steps: (1) deduplication of data, (2) deletion of intersection, and (3) an exhaustive search. The best mating (BM) 25 and length normalization of BM25 (BM25L) algorithms were used to construct positive and negative pairs. The performance of the model was evaluated using the 5-fold cross-validation ensemble technique. Experiments demonstrate the effectiveness of the proposed method quantitatively. Moreover, we expect this method to be optimal for increasing software productivity in various NLP tasks
Sentiment-Target Word Pair Extraction Model Using Statistical Analysis of Sentence Structures
No full textProduct information has been propagated online via forums and social media. Lots of merchandise are recommended via an expert system method and is considered for purchase by online comments or product reviews. For predicting people’s opinions on products, studying people’s thoughts via extracting information in documents is referred to as sentiment analysis. Finding sentiment-target word pairs is an important sentiment mining research issue. With the Korean language, as the predicate appears at the very end, it is not easy to find the exact word pairs without first identifying the syntactic structure of the sentence. In this study, we propose a model that parses sentence structures and extracts sentiment-target word pairs from the parse tree. The proposed model extracts the sentiment-target word pairs that appear in the sentence by using parsing and statistical methods. For extracting sentiment-target word pairs, this model uses a sentiment word extractor and a target word extractor. After testing data from 4000 movie reviews, the applicable model showed high performance in both accuracy 93.25 (+14.45) and F1-score 82.29 (+3.31) compared with others. However, improvements in the recall rate (−0.35) are needed and computational costs must be reduced
A Study on the Development of Game-based Mind Wandering Judgment Model in Video Lecture-based Education
Full text linkAlthough video lecture materials are very efficient learning materials, they are likely to be unilateral learning materials by the lecturer. It is easily degraded to be one-sided learning, which has been considered as a problem of online education, and it is difficult to judge whether learners are actually learning. Therefore, in this paper, a minimum learning activity judgment model that can automatically determine if they actually learn through mind wandering judgment was proposed to overcome the limitations of previous learning materials, and educational effect verification experiment was performed. Experiment results show that the video lecture class using the minimum learning activity judgment system was effective in improving the academic achievement
Automatic Judgement of Online Video Watching: I Know Whether or Not You Watched
No full textVideos have long been viewed through the free choice of customers, but in some cases currently, watching them is absolutely required, for example, in institutions, companies, and education, even if the viewers prefer otherwise. In such cases, the video provider wants to determine whether the viewer has honestly been watching, but the current video viewing judging system has many loopholes; thus, it is hard to distinguish between honest viewers and false viewers. Time interval different answer popup quiz (TIDAPQ) was developed to judge honest watching. In this study, TIDAPQ randomly inserts specially developed popup quizzes in the video. Viewers must solve time interval pass (RESULT 1) and individually different correct answers (RESULT 2) while they watch. Then, using these two factors, TIDAPQ immediately performs a comprehensive judgement on whether the viewer honestly watched the video. To measure the performance of TIDAPQ, 100 experimental subjects were recruited to participate in the model verification experiment. The judgement performance on normal watching was 93.31%, and the judgement performance on unusual watching was 85.71%. We hope this study will be useful in many areas where watching judgements are needed
Unifying user preference and item knowledge-based similarity models for top-N recommendation
No full text
