• ETRI Journal
• /
• v.46 no.1
• /
• pp.59-70
• /
• 2024

We introduce a high-performance named entity recognition (NER) model for written and spoken language. To overcome challenges related to labeled data scarcity and domain shifts, we use transfer learning to leverage our previously developed KorBERT as the base model. We also adopt a meta-pseudo-label method using a teacher/student framework with labeled and unlabeled data. Our model presents two modifications. First, the student model is updated with an average loss from both human- and pseudo-labeled data. Second, the influence of noisy pseudo-labeled data is mitigated by considering feedback scores and updating the teacher model only when below a threshold (0.0005). We achieve the target NER performance in the spoken language domain and improve that in the written language domain by proposing a straightforward rollback method that reverts to the best model based on scarce human-labeled data. Further improvement is achieved by adjusting the label vector weights in the named entity dictionary.

• The Mathematical Education
• /
• v.31 no.2
• /
• pp.93-107
• /
• 1992

The purpose of this study is to make out teaching-learning method for developing mathematical abilities of the 1st grade children in elementary school by investigating cognitive effects which mathematical pre-experiences given intentionally by teachers have on children's learning mathematics. The research questions for this purpose are as follows: In learning effects through mathematical pre-experiences given intentionally by teachers. 1) is there any differences between children with pre-experiences and children without them in Mathematics Achievement Test\ulcorner 2) is there any differences between children with pre-experiences and children without them in Transfer Test for learning effects\ulcorner For this study, a class with 41 children in H elementary school located in a Myon near Chong-ju was selected as an experimental group and a class with 43 children in G elementary school in the same Myon was selected as a control group. Nonequivalent Control Group Design of Quasi-Experimental Design was applied to this study. To give pre-experiences to the children in experimental group, their classroom was equipped with materials for pre-experiences, so children could always observe the materials and play with them. The materials were a round-clock on the wall, two pairs of scales, fifty dice, some small pebbles, two pairs of weight scales, two rulers on the wall, and various cards for playing games. Pre-experiences were given to the children repeatedly through games and observations during free time in the morning (00:20-09:00) and intervals between periods. There was a pretest for homogeneity of mathematics achievement between the two groups and were Mathematics Achievement Test (30 items) and Transfer Test (25 items) for learning effects as post-tests. The data were collected from the pretest on April 8 (control group), on April 11 (experimental group) and from the Mathematics Achievement Test and Transfer Test on July 15 (experimental group) and on July 16 (control group). T-test was used to analyze if there were any differences in the results of the test. The results of the analysis were as follows: (1) As the result of pretest, there was not a significance difference between the experimental group (M=17.10. SD=7.465) and the control group (M=16.31, SD=6.974) at p<.05 (p=0.632). (2) For the question 1. in the Mathematics Achievement Test, there was a significant difference between the experimental group (M=26.08, SD=4.827) and the control group (M=22.28. SD=5.913) at p<.01 (p=.003). (3) For the question 2. in the Transfer Test for learning effects. there was a significant difference between the experimental group (M=16.41, SD=5.800) and the control group (M=11.84, SD=4.815) at p<001, (p=.000). From the results of the analyses obtained in this study. the following conclusions can be drawn: First, mathematical pre-experiences given by teachers are effective in increasing mathematical achievement and transfer in learning mathematics. Second, games. observations, and experiments given intentionally by teachers can make children's mathematical experiences rich and various, and are effective in adjusting individual differences for the mathematical experiences obtained before they entered elementary schools. Third, it is necessary for teachers to give mathematical pre-experiences with close attention in order to stimulate children's mathematical interests and intellectual curiosity.

• Journal of Industrial Convergence
• /
• v.20 no.2
• /
• pp.61-70
• /
• 2022

The study developed a psychiatric and mental health nursing practice program based on blended learning as nursing students' field practice in psychiatric and mental nursing practice was limited due to the prolonged COVID-19 pandemic. This is a study to evaluate the effect on communication competency, self-efficacy for group work, and learning transfer motivation through a psychiatric and mental health nursing practice program based on blended learning. From October 18, 2021 to December 11, 2021, 64 nursing students participated in the study using a structured Google questionnaire. The collected data was analyzed by descriptive statistics and paired t-test using the SPSS 25.0 program. As a result of the study, based on blended learning, the subjects' communication competency, self-efficacy for group work, and learning transfer motivation were significantly increased after compared to before psychiatric and mental health nursing practice education. Through the results of this study, it was possible to confirm the effect of the psychiatric and mental health nursing practice program based on blended learning.

• Knowledge Management Research
• /
• v.11 no.4
• /
• pp.41-66
• /
• 2010

This research examines the effect of the relatedness and the gap between Resources and mechanisms on effectiveness of inter-organizational knowledge transfer. According to the literature, there has been a competing theory between two claims; one is that inter-organizational knowledge transfer will be more effective due to the reduction of the transaction cost as the relatedness increases. And the other is that the mutual complementarity of different organizational characteristics will increase synergy. In total, the relatedness and the gap of the Resource and mechanism makes the inverted U-shaped relationship with the inter-organizational knowledge transfer. As the result of empirical analysis about 109 Korean-based Joint Ventures entered country, it shows that the relatedness of parent company's production Resources, learning mechanisms, and coordination mechanisms made the inverted U-shaped relations with the inter-organizational knowledge transfer and the gap of production Resources and adjustment mechanism formed the same relationship. However, the U-shaped relationship has been established in the relatedness of market Resources, but the gap of market Resources and the learning mechanism was not statistically significant. Through this study, I can draw a best conclusion that the inter-organizational knowledge transfer will be more effective when the relatedness and the gap of management resources and mechanisms is in optimal level. However, when it comes to market Resources, it can be inferred that the result could be the opposite because the partner country's market environment would be different.

• Journal of Agricultural Extension & Community Development
• /
• v.4 no.1
• /
• pp.305-315
• /
• 1997

The propose of this study was to analyze the relationships among the levels of training & development evaluation (reaction, learning, transfer). The study has been conducted on 730 trainees who attended in the basic accounting program in L training and development institution through three incidents of tracked research such as reaction survey right after the conclusion of training, learning evaluation through test, and an evaluation of the transferability after 3 months of training. Questionnaires and test papers for analyses were used after their reliability, validity, difficulty, and discrimination have been verified on a pre-test. The research has been conducted for six months from 4 March 1996 to the end of August 1996, and data have been collected through direct research and survey through mail. The collected data have been worked on at SAS program for Windows with a statistical significance level of 5%. Statistical method that had been used was Pearson's correlation coefficient. The result and conclusion acquired from this study were as follows: Between reaction and learning, learning and transfer of training, only a weak positive correlation exists and explanation or prediction variance showing hierarchical relationship was quite weak with 1%. Thus, this research not only does not strongly support Kirkpatrick(1976)'s hierarchical model of $reaction{\rightarrow}learning{\rightarrow}transfer$, but also indicates that the separate measurement on each levels of training evaluation needs to be done. On the other hand, there was a relatively strong positive correlation between reaction and transfer of training. Based on the result, the conclusion, and the restriction perceived through this study, the following suggestions were made. 1. There is a need to empirically analyze and verify the hierarchy of all levels of training evaluation including the evaluation of the fourth level (result) such as organizational productivity, organizational satisfaction, and separation rate. 2. A great deal of efforts will be needed to systematically analyze what the relationships are among the methods measuring the level of evaluation of the training and development, and to apply this result to the training field.

• Imaging Science in Dentistry
• /
• v.52 no.2
• /
• pp.219-224
• /
• 2022

Purpose: This study aimed to evaluate the performance of transfer learning in a deep convolutional neural network for classifying implant fixtures. Materials and Methods: Periapical radiographs of implant fixtures obtained using the Superline (Dentium Co. Ltd., Seoul, Korea), TS III(Osstem Implant Co. Ltd., Seoul, Korea), and Bone Level Implant(Institut Straumann AG, Basel, Switzerland) systems were selected from patients who underwent dental implant treatment. All 355 implant fixtures comprised the total dataset and were annotated with the name of the system. The total dataset was split into a training dataset and a test dataset at a ratio of 8 to 2, respectively. YOLOv3 (You Only Look Once version 3, available at https://pjreddie.com/darknet/yolo/), a deep convolutional neural network that has been pretrained with a large image dataset of objects, was used to train the model to classify fixtures in periapical images, in a process called transfer learning. This network was trained with the training dataset for 100, 200, and 300 epochs. Using the test dataset, the performance of the network was evaluated in terms of sensitivity, specificity, and accuracy. Results: When YOLOv3 was trained for 200 epochs, the sensitivity, specificity, accuracy, and confidence score were the highest for all systems, with overall results of 94.4%, 97.9%, 96.7%, and 0.75, respectively. The network showed the best performance in classifying Bone Level Implant fixtures, with 100.0% sensitivity, specificity, and accuracy. Conclusion: Through transfer learning, high performance could be achieved with YOLOv3, even using a small amount of data.

• Journal of Information Processing Systems
• /
• v.18 no.4
• /
• pp.535-548
• /
• 2022

Using a rich resource language to classify sentiments in a language with few resources is a popular subject of research in natural language processing. Burmese is a low-resource language. In light of the scarcity of labeled training data for sentiment classification in Burmese, in this study, we propose a method of transfer learning for sentiment analysis of a language that uses the feature transfer technique on sentiments in English. This method generates a cross-language word-embedding representation of Burmese vocabulary to map Burmese text to the semantic space of English text. A model to classify sentiments in English is then pre-trained using a convolutional neural network and an attention mechanism, where the network shares the model for sentiment analysis of English. The parameters of the network layer are used to learn the cross-language features of the sentiments, which are then transferred to the model to classify sentiments in Burmese. Finally, the model was tuned using the labeled Burmese data. The results of the experiments show that the proposed method can significantly improve the classification of sentiments in Burmese compared to a model trained using only a Burmese corpus.

• The Journal of the Convergence on Culture Technology
• /
• v.10 no.3
• /
• pp.801-807
• /
• 2024

Linear motion robots are devices that perform functions such as transferring parts or positioning devices, and require high precision. In companies that develop linear robot application systems, human workers are in charge of quality control and fault diagnosis of linear robots, and the result and accuracy of a fault diagnosis varies depending on the skill level of the person in charge. Recently, there have been many attempts to utilize artificial intelligence to diagnose faults in industrial devices. In this paper, we present a system that automatically diagnoses linear rail and ball screw misalignment of a linear robot using transfer learning. In industrial systems, it is difficult to obtain a lot of learning data, and this causes a data imbalance problem. In this case, a transfer learning model configured by retraining an established model is widely used. The information obtained by using an acceleration sensor and torque sensor was used, and its usefulness was evaluated for each case. After converting the signal obtained from the sensor into a spectrogram image, the type of abnormality was diagnosed using an image recognition artificial intelligence classifier. It is expected that the proposed method can be used not only for linear robots but also for diagnosing other industrial robots.

• Journal of KIISE
• /
• v.44 no.12
• /
• pp.1275-1281
• /
• 2017

The supervised learning approach is suitable for classification of insulting sentences, but pre-decided training sentences are necessary. Since a Character-level Convolution Neural Network is robust for each character, so is appropriate for classifying abusive sentences, however, has a drawback that demanding a lot of training sentences. In this paper, we propose transfer learning method that reusing the trained filters in the real classification process after the filters get the characteristics of offensive words by generated abusive/normal pair of sentences. We got higher performances of the classifier by decreasing the effects of data shortage and class imbalance. We executed experiments and evaluations for three datasets and got higher F1-score of character-level CNN classifier when applying transfer learning in all datasets.

• Information Systems Review
• /
• v.12 no.3
• /
• pp.19-48
• /
• 2010

This study focused on characterizing various factors in order for learners to maintain their interests in learning and to maximize learning effects as the top priority purpose of university e-Learning, on the basis of results of conceptual studies on existing e-Learning and practical studies, and then on examining them practically. It also analyzed which factors would have greater influence on learning effects of e-Learning in general. Moreover, in comparison with existing numerous studies which examined only factor such as learning effects of e-Learning, it analyzed such things in detail according to division into three items such as learning satisfaction, learning transfer and learning recommendation. To achieve such purposes of the study, it characterized and set 3 factors such as learning contents, instructional design and user convenience on the assumption that such factors have a significant influence on learning effects of e-Learning. Moreover, the factor of learning contents includes 3 detailed elements, i.e., learning issue and objective, knowledge information, and consistency and propriety, and the factor of instructional design includes 4 detailed elements, i.e., interest and sympathy, interaction, contents presentation and explanatory strategy. Lastly, the factor of user convenience includes 2 detailed elements such as screen configuration, and check-up of contents and teaching schedule. According to analytical results, it showed all 3 factors such as learning contents, instructional design and user convenience have a significant influence on learning effects of e-Learning(i.e., learning satisfaction, learning transfer and learning recommendation). In more detail, it showed the learning issue and objective from the factor of learning contents have the greatest influence on learning satisfaction of e-Learning. Then, it is the most important to set the learning issue and objective with given priority to learners and set the learning objective estimable, in order to raise the learning satisfaction. It showed the contents presentation from the factor of instructional design on the learning transfer. Therefore, it is the most important to structuralize mutual relation and presentation orders to promote learning systematically and to let learners access to such things, for the purpose of raising the learning transfer. Moreover, it showed the interest and sympathy from the factor of instructional design has the greatest influence on the learning recommendation. Thus, it is the most important to promote learners' interests to the maximum using well-timed media, and to give a lecture enough to arouse learners' sympathy.