• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.19 no.11
• /
• pp.310-318
• /
• 2018

Recently, the importance of big data analysis is increasing as a large amount of data is generated by various devices connected to the Internet with the advent of Internet of Things (IoT). Especially, it is necessary to analyze various large-scale IoT streaming sensor data generated in real time and provide various services through new meaningful prediction. This paper proposes a real-time indoor PM10 concentration prediction LSTM model based on streaming data generated from IoT sensor using AWS. We also construct a real-time indoor PM10 concentration prediction service based on the proposed model. Data used in the paper is streaming data collected from the PM10 IoT sensor for 24 hours. This time series data is converted into sequence data consisting of 30 consecutive values from time series data for use as input data of LSTM. The LSTM model is learned through a sliding window process of moving to the immediately adjacent dataset. In order to improve the performance of the model, incremental learning method is applied to the streaming data collected every 24 hours. The linear regression and recurrent neural networks (RNN) models are compared to evaluate the performance of LSTM model. Experimental results show that the proposed LSTM prediction model has 700% improvement over linear regression and 140% improvement over RNN model for its performance level.

• The Journal of Korean society of community based occupational therapy
• /
• v.8 no.2
• /
• pp.15-28
• /
• 2018

Objective : This study sought to characterize and determine the qualitative level of a single target study for stroke patients. Methods : The National Science and Technology Information Center (NDSL), DBpia (DBpia), RISS (Radical Research Information Service), Korea Research Information (KISS), and the National Assembly Library's original case study from 2002 to 2017. A total of 24 single target research papers were selected through the screening process to analyze the quality level of research methods and research design. Results : ABA design was th most common study design method. One person was the most with 12(50%). and three were the second with 8(33.3%). Imagination was the most used as an independent lawyer. Dependent variables had the highest level of situability and one-sidedness. The study was also conducted with a variety of target behaviors, including 'memory', 'visual attention', 'dysphagia', 'visual-motor coordination', 'balance', 'activity of daily life' and 'edema' behaviors. It also showed a positive effect on all dependent variables. The Qualitative level was found to be above the intermediate level except for one study. Conclusion : It is academic significance that this study analyzes the items to be prepared for in the performance of a single target study and further studies may require the establishment of a weak but good-quality single target study for researchers conducting research in local communities and clinical sites.

• The Journal of the Korea Contents Association
• /
• v.19 no.9
• /
• pp.308-320
• /
• 2019

This study aims at investigating the major symptoms that help family carers detect mental illness in elderly patients. Another purpose of this study is to empirically verify the major factors determining the utilization of mental health services with a focus on family carers. The results of this study are as follows. First, the most commonly detected symptoms that caused the family carers to suspect mental illness in the elderly patients were memory decline and other forms of cognitive function decline. Second, the determinants of the elderly's utilization of mental health services included the patient's long-term care insurance level, the age of the family carer, the period of care, the level stress associated with the provision of care felt by the carer, his understanding of geriatric mental illness, and the level of perception about community mental health services. Based on these findings, this study suggests policies and practical implications for the early detection of and response to elderly mental health problems and the utilization of related services from the viewpoint of the family carers of the elderly.

• The Korean Journal of Archival Studies
• /
• no.48
• /
• pp.141-179
• /
• 2016

In May 2011, the 1980 Archives for the May 18th Democratic Uprising or the "5 18 Gwangju Democratization Movement" were registered as a UNESCO Memory of the World. This historic moment told the world that the 5 18 movement is a valuable and historical asset. However, despite the international recognition, archives that deny of such facts are still rampant because of misdirected standings and prejudices. These sources even develop discourses by distorting the archives to justify their claims. Accordingly, the study aimed to identify how these sources form ideologies surrounding the 5 18 movement discourses, which are characteristic of extreme social standings. It explored the possibility of the distortion of archives presented for each discourse and reconsidered the archivists' positions and roles to cope with such possibility. In addition, the author aims to suggest a more systemic strategy to advance the existing responses against the distortions, as well as provide discourses that are based on true and accurate archives mainly to students who have not yet been introduced to such distorted discourses. In the future, archivists shall try to develop positive awareness about the 5 18 discourses rather than maintain passive positions that provide information from limited archives. Through this, it is expected that this study will advance future analyses that would be effective against the distortion of archives.

• Management & Information Systems Review
• /
• v.37 no.4
• /
• pp.237-256
• /
• 2018

This study focused on the concept of organizational learning as a prior variable of innovation activities, and reviewed the relationship between organizational learning, innovation and management performance. According to prior studies, the ability to perform these activities may be enhanced through organizational learning, as the success of the innovation requires activities to acquire and share knowledge within the organization. In other words, organizational learning is playing a role as a precursor to innovation. Therefore, in this study, the effects of organizational learning on management performance are to be verified through the mediation effect of product and innovation activities. Organizational learning provides various definitions and components for each scholar, but this study consisted of a series of knowledge acquisition, information distribution, information analysis and process memory using the framework of the learning ability analysis by Levitt and March(1988) and Huber(1991), Innovation was also divided into product innovation and process innovation, and measured with sub-variables such as presentation of new products and improvement activities to increase productivity. Management performance was measured as financial and non-financial performance. To verify the effects of the mediation, we used a three-step regression analysis procedure of Baron and Kenny(1986)'s and a sobel-test. Empirical studies show that organizational learning has a positive effect on management performance and that knowledge acquisition and information distribution, which are the early stages of learning activities in the lower variables, affect performance through product innovation. Based on the results of the above empirical study, the implications, limitations of the study and future research directions were presented.

• KIPS Transactions on Software and Data Engineering
• /
• v.10 no.5
• /
• pp.195-200
• /
• 2021

Due to the development of video sharing platforms, the amount of video uploads is exploding. Such videos often include various types of music, among which cover songs are included. In order to protect the copyright of music, an algorithm to find the original song of the cover song is essential. However, it is not easy to find the original song because the cover song is a modification of the composition, speed and overall structure of the original song. So far, there is no known effective algorithm for searching the original song of the cover song. In this paper, we propose an algorithm for searching the original song of the cover song using the inflection points of the melody line. Inflection points represent the characteristic points of change in the melody sequence. The proposed algorithm compares the original song and the cover song using the sequence of inflection points for the representative phrase of the original song. Since the characteristics of the representative phrase are used, even if the cover song is a song made by modifying the overall composition of the song, the algorithm's search performance is excellent. Also, since the proposed algorithm uses only the features of the inflection point sequence, the memory usage is very low. The efficiency of the algorithm was verified through performance evaluation.

• Journal of Digital Convergence
• /
• v.17 no.11
• /
• pp.21-27
• /
• 2019

The study empirically analyzed the effects of content attitudes and recall on digital signage advertising appeal (information appeal vs. image appeal) and interactivity level (low vs. high). As a result, first, it was found that a moderately low level of interactivity had a positive effect on content attitudes and recall than when the level of digital signage was extremely high. In addition, at moderately low levels of interactivity, information appeals had higher content attitudes and recalls than image appeals. Second, the content of image appeal has a positive effect on attitude when the digital signage level of interactivity is high, and the image recall ad and information appeal ad have negative effects on recall. Third, the low level of interactivity of digital signage has a positive effect on the content attitude and recall of information appeal. With the advent of digital media in recent years, concerns about how to construct the level of interactivity and information content on a strategic level are increasing in practice. The results of this study are expected to suggest the direction of the strategic grounds for this.

• The Journal of Korean Philosophical History
• /
• no.59
• /
• pp.343-370
• /
• 2018

Jeong Si-Han(丁時翰) completed "Siqibianzheng"("四七辨證") at the age of 72, and later had the argument over Runwuxingtongyilunzheng(人物性同異論爭) with his disciple, Lee Sik(李?). Jeong Si-Han(丁時翰) had the position of Runwuxingyilun(人物性異論) and Lee Sik(李?) Runwuxingtonglun(人物性同論). Yet, the argument over Runwuxingtongyilunzheng(人物性同異論爭) had been forgotten and "Siqibianzheng"("四七辨證") could be acknowledged because Toegye school(退溪學派) and Yulgok school(栗谷學派) were conflicting and criticizing each other's stance at that time. It seems like Lee Hyeon-Il(李玄逸)'s "Liqulishilunsiduanqiqingshubian"("栗谷李氏論四端七情書辨") had a great influence on the completion of "Siqibianzheng" ("四七辨證"). Lee Yi(李珥)'s thought of Siduan(四端) and Qiqing(七情) was 'the position of Hunlun(渾淪)' that 'Qiqing(七情) includes Siduan(四端)', and Lee Hwang(李滉) had the 'position of Fenkai(分開)' that Siduan(四端) and Qiqing(七情) should be interpreted differently. Jeong Si-Han(丁時翰) criticized the stance of Hunlun(渾淪) from the position of Fenkai(分開). What did Jeong Si-Han(丁時翰) try to pursue through "Siqibianzheng"("四七辨證")? This fact tends to make us forget the controversy over Runwuxingtongyilunzheng(人物性同異論爭) between Jeong Si-Han(丁時翰) and disciple Lee Sik(李?). Now we know the fact that Jeong Si-Han(丁時翰) criticized Lee Yi(李珥) in "Siqibianzheng"("四七辨證"), but don't care much about the fact that Jeong Si-Han(丁時翰) caused the controversy over Runwuxingtongyilun(人物性同異論). Why "Siqibianzheng"("四七辨證") has remained in our memory even though it was an important one? It might be because "Siqibianzheng"("四七辨證") had an impact on Lee Sik(李?), Shin Hu-Dam(愼後聃), and even Jeong Yak-Yong(丁若鏞) in the process of summarizing the arguments over Siduanqiqinglunzheng(四端七情論爭) since Jeong Si-Han(丁時翰) regardless of whether Jeong Si-Han(丁時翰) had hoped it or not.

• The Journal of the Convergence on Culture Technology
• /
• v.7 no.1
• /
• pp.640-645
• /
• 2021

The deep learning based end-to-end TTS system consists of Text2Mel module that generates spectrogram from text, and vocoder module that synthesizes speech signals from spectrogram. Recently, by applying deep learning technology to the TTS system the intelligibility and naturalness of the synthesized speech is as improved as human vocalization. However, it has the disadvantage that the inference speed for synthesizing speech is very slow compared to the conventional method. The inference speed can be improved by applying the non-autoregressive method which can generate speech samples in parallel independent of previously generated samples. In this paper, we introduce FastSpeech, FastSpeech 2, and FastPitch as Text2Mel technology, and Parallel WaveGAN, Multi-band MelGAN, and WaveGlow as vocoder technology applying non-autoregressive method. And we implement them to verify whether it can be processed in real time. Experimental results show that by the obtained RTF all the presented methods are sufficiently capable of real-time processing. And it can be seen that the size of the learned model is about tens to hundreds of megabytes except WaveGlow, and it can be applied to the embedded environment where the memory is limited.

• KIPS Transactions on Computer and Communication Systems
• /
• v.10 no.3
• /
• pp.71-80
• /
• 2021

The selection of an appropriate neural network algorithm is an important step for accurate data prediction in machine learning. Many algorithms based on basic artificial neural networks have been devised to efficiently predict future data. These networks include deep neural networks (DNNs), recurrent neural networks (RNNs), long short-term memory (LSTM) networks, and gated recurrent unit (GRU) neural networks. Developers face difficulties when choosing among these networks because sufficient information on their performance is unavailable. To alleviate this difficulty, we evaluated the performance of each algorithm by comparing their errors and processing times. Each neural network model was trained using a tax dataset, and the trained model was used for data prediction to compare accuracies among the various algorithms. Furthermore, the effects of activation functions and various optimizers on the performance of the models were analyzed The experimental results show that the GRU and LSTM algorithms yields the lowest prediction error with an average RMSE of 0.12 and an average R2 score of 0.78 and 0.75 respectively, and the basic DNN model achieves the lowest processing time but highest average RMSE of 0.163. Furthermore, the Adam optimizer yields the best performance (with DNN, GRU, and LSTM) in terms of error and the worst performance in terms of processing time. The findings of this study are thus expected to be useful for scientists and developers.