• Proceedings of the KSPS conference
• /
• 2007.05a
• /
• pp.207-210
• /
• 2007

As the Internet is prevalent in our life, harmful contents have been increasing on the Internet, which has become a very serious problem. Among them, pornographic video is harmful as poison to our children. To prevent such an event, there are many filtering systems which are based on the keyword based methods or image based methods. The main purpose of this paper is to devise a system that classifies the pornographic videos based on the audio information. We use Mel-Cepstrum Modulation Energy (MCME) which is modulation energy calculated on the time trajectory of the Mel-Frequency cepstral coefficients (MFCC) and MFCC as the feature vector and Gaussian Mixture Model (GMM) as the classifier. With the experiments, the proposed system classified the 97.5% of pornographic data and 99.5% of non-pornographic data. We expect the proposed method can be used as a component of the more accurate classification system which uses video information and audio information simultaneously.

• Journal of the Korean Institute of Navigation
• /
• v.23 no.1
• /
• pp.75-84
• /
• 1999

Although many studies on modelling of port competitive situation have been conducted, both theoretical frame and methodology are still very weak. In this study, therefore, a new algorithm called ESD (Extensional System Dynamics) for the evaluation of port competition was presented, and applied to simulate port systems in northeast asia. The detailed objectives of this paper are to develop Unit fort Model by using SD(System Dynamics) method; to develop Competitive Port Model by ESD method; to perform sensitivity analysis by altering parameters, and to propose port development strategies. For these the algorithm for the evaluation of part's competition was developed in two steps. Firstly, SD method was adopted to develop the Unit Port models, and secondly HFP(Hierarchical Fuzzy Process) method was introduced to expand previous SD method. The proposed models were then developed and applied to the five ports - Pusan, Kobe, Yokohama, Kaoshiung, Keelung - with real data on each ports, and several findings were derived. Firstly, the extraction of factors for Unit Port was accomplished by consultation of experts such as research worker, professor, research fellows related to harbor, and expert group, and finally, five factor groups - location, facility, service, cargo volumes, and port charge - were obtained. Secondly, system's structure consisting of feedback loop was found easily by location of representative and detailed factors on keyword network of STGB map. Using these keyword network, feedback loop was found. Thirdly, for the target year of 2003, the simulation for Pusan port revealed that liner's number would be increased from 829 ships to 1,450 ships and container cargo volumes increased from 4.56 million TEU to 7.74 million TEU. It also revealed that because of increased liners and container cargo volumes, length of berth should be expanded from 2,162m to 4,729m. This berth expansion was resulted in the decrease of congested ship's number from 97 to 11. It was also found that port's charge had a fluctuation. Results of simulation for Kobe, Yokohama, Kaoshiung, Keelung in northeast asia were also acquired. Finally, the inter port competition models developed by ESB method were used to simulate container cargo volumes for Pusan port. The results revealed that under competitive situation container cargo volume was smaller than non-competitive situation, which means Pusan port is lack of competitive power to other ports. Developed models in this study were then applied to estimate change of container cargo volumes in competitive relation by altering several parameters. And, the results were found to be very helpful for port mangers who are in charge of planning of port development.

• Review of Korean Society for Internet Information
• /
• v.14 no.3
• /
• pp.68-77
• /
• 2013

This article consists of 3 parts. Part I is multi stakeholder approach on Internet governance system. Part II is analysis of the Korean Internet governance system. In this part, I explain relevant laws in Korea, including Korean Internet Address Resources Act. Part III is my suggestion on Korean Internet governance system using a multi stakeholder approach. First of all, the keyword of the Internet governance system is decision making process: that is, consensus based versus top-down approach. Then who are major players in Internet governance in national level? Government, or Private sectors such as business and civil society. Korean legal system for Internet governance shows a top-down decision making process. Major players are the government (that is, Ministry of Science, ICT and Future Planning) and KISA affiliated with the government. Other players include Internet Address Policy Committee, Korea Internet Governance Alliance, and NGOs. The key statute for Internet governance in Korea is Internet Address Resources Act of 2004. Articles 3 and 5 require the Ministry of Science, ICT and Future Planning to take a proactive role in Internet governance. The government shall consult with the Internet Address Policy Deliberation Committee for Internet governance. Yet this Committee is established under the control of the Ministry of Science, ICT and Future Planning. All members of this Committee are also commissioned or nominated by the Chairman of the Ministry. Meanwhile, there are also non-official organizations, including Sub-committee on Address & Infrastructure of Korea Internet Governance Alliance. I suggest to reform decision making process of Korean Internet governance system based on BOTTOM-UP process for CONSENSUS BASED DECISION. My suggested system includes the following: (1) The government hands over a major role in Internet governance to INDEPENDENT Internet policy organization. And the government participates in such organization as ONE of the players. (2) Nomination of this committee member must be bottom-up process for a genuine multi-stakeholder model including civil society, commercial organization, end-users and experts. (3) The government should establish plan for supporting the private sector's international activity on the long-term basis.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.11 no.5
• /
• pp.467-474
• /
• 2018

To organize and document the contents of meetings and discussions is very important in various tasks. However, in the past, people had to manually organize the contents themselves. In this paper, we describe the development of a system that generates the meeting minutes automatically using the TextRank algorithm. The proposed system records all the utterances of the speaker in real time and calculates the similarity based on the appearance frequency of the sentences. Then, to create the meeting minutes, it extracts important words or phrases through a non-supervised learning algorithm for finding the relation between the sentences in the document data. Especially, we improved the performance by introducing the keyword weighting technique for the TextRank algorithm which reconfigured the PageRank algorithm to fit words and sentences.

• Journal of Intelligence and Information Systems
• /
• v.27 no.3
• /
• pp.175-197
• /
• 2021

Recently, with the development of deep learning technology, research on unstructured data analysis is being actively conducted, and it is showing remarkable results in various fields such as classification, summary, and generation. Among various text analysis fields, text classification is the most widely used technology in academia and industry. Text classification includes binary class classification with one label among two classes, multi-class classification with one label among several classes, and multi-label classification with multiple labels among several classes. In particular, multi-label classification requires a different training method from binary class classification and multi-class classification because of the characteristic of having multiple labels. In addition, since the number of labels to be predicted increases as the number of labels and classes increases, there is a limitation in that performance improvement is difficult due to an increase in prediction difficulty. To overcome these limitations, (i) compressing the initially given high-dimensional label space into a low-dimensional latent label space, (ii) after performing training to predict the compressed label, (iii) restoring the predicted label to the high-dimensional original label space, research on label embedding is being actively conducted. Typical label embedding techniques include Principal Label Space Transformation (PLST), Multi-Label Classification via Boolean Matrix Decomposition (MLC-BMaD), and Bayesian Multi-Label Compressed Sensing (BML-CS). However, since these techniques consider only the linear relationship between labels or compress the labels by random transformation, it is difficult to understand the non-linear relationship between labels, so there is a limitation in that it is not possible to create a latent label space sufficiently containing the information of the original label. Recently, there have been increasing attempts to improve performance by applying deep learning technology to label embedding. Label embedding using an autoencoder, a deep learning model that is effective for data compression and restoration, is representative. However, the traditional autoencoder-based label embedding has a limitation in that a large amount of information loss occurs when compressing a high-dimensional label space having a myriad of classes into a low-dimensional latent label space. This can be found in the gradient loss problem that occurs in the backpropagation process of learning. To solve this problem, skip connection was devised, and by adding the input of the layer to the output to prevent gradient loss during backpropagation, efficient learning is possible even when the layer is deep. Skip connection is mainly used for image feature extraction in convolutional neural networks, but studies using skip connection in autoencoder or label embedding process are still lacking. Therefore, in this study, we propose an autoencoder-based label embedding methodology in which skip connections are added to each of the encoder and decoder to form a low-dimensional latent label space that reflects the information of the high-dimensional label space well. In addition, the proposed methodology was applied to actual paper keywords to derive the high-dimensional keyword label space and the low-dimensional latent label space. Using this, we conducted an experiment to predict the compressed keyword vector existing in the latent label space from the paper abstract and to evaluate the multi-label classification by restoring the predicted keyword vector back to the original label space. As a result, the accuracy, precision, recall, and F1 score used as performance indicators showed far superior performance in multi-label classification based on the proposed methodology compared to traditional multi-label classification methods. This can be seen that the low-dimensional latent label space derived through the proposed methodology well reflected the information of the high-dimensional label space, which ultimately led to the improvement of the performance of the multi-label classification itself. In addition, the utility of the proposed methodology was identified by comparing the performance of the proposed methodology according to the domain characteristics and the number of dimensions of the latent label space.

• Journal of Intelligence and Information Systems
• /
• v.19 no.3
• /
• pp.1-23
• /
• 2013

To discover significant social issues such as unemployment, economy crisis, social welfare etc. that are urgent issues to be solved in a modern society, in the existing approach, researchers usually collect opinions from professional experts and scholars through either online or offline surveys. However, such a method does not seem to be effective from time to time. As usual, due to the problem of expense, a large number of survey replies are seldom gathered. In some cases, it is also hard to find out professional persons dealing with specific social issues. Thus, the sample set is often small and may have some bias. Furthermore, regarding a social issue, several experts may make totally different conclusions because each expert has his subjective point of view and different background. In this case, it is considerably hard to figure out what current social issues are and which social issues are really important. To surmount the shortcomings of the current approach, in this paper, we develop a prototype system that semi-automatically detects social issue keywords representing social issues and problems from about 1.3 million news articles issued by about 10 major domestic presses in Korea from June 2009 until July 2012. Our proposed system consists of (1) collecting and extracting texts from the collected news articles, (2) identifying only news articles related to social issues, (3) analyzing the lexical items of Korean sentences, (4) finding a set of topics regarding social keywords over time based on probabilistic topic modeling, (5) matching relevant paragraphs to a given topic, and (6) visualizing social keywords for easy understanding. In particular, we propose a novel matching algorithm relying on generative models. The goal of our proposed matching algorithm is to best match paragraphs to each topic. Technically, using a topic model such as Latent Dirichlet Allocation (LDA), we can obtain a set of topics, each of which has relevant terms and their probability values. In our problem, given a set of text documents (e.g., news articles), LDA shows a set of topic clusters, and then each topic cluster is labeled by human annotators, where each topic label stands for a social keyword. For example, suppose there is a topic (e.g., Topic1 = {(unemployment, 0.4), (layoff, 0.3), (business, 0.3)}) and then a human annotator labels "Unemployment Problem" on Topic1. In this example, it is non-trivial to understand what happened to the unemployment problem in our society. In other words, taking a look at only social keywords, we have no idea of the detailed events occurring in our society. To tackle this matter, we develop the matching algorithm that computes the probability value of a paragraph given a topic, relying on (i) topic terms and (ii) their probability values. For instance, given a set of text documents, we segment each text document to paragraphs. In the meantime, using LDA, we can extract a set of topics from the text documents. Based on our matching process, each paragraph is assigned to a topic, indicating that the paragraph best matches the topic. Finally, each topic has several best matched paragraphs. Furthermore, assuming there are a topic (e.g., Unemployment Problem) and the best matched paragraph (e.g., Up to 300 workers lost their jobs in XXX company at Seoul). In this case, we can grasp the detailed information of the social keyword such as "300 workers", "unemployment", "XXX company", and "Seoul". In addition, our system visualizes social keywords over time. Therefore, through our matching process and keyword visualization, most researchers will be able to detect social issues easily and quickly. Through this prototype system, we have detected various social issues appearing in our society and also showed effectiveness of our proposed methods according to our experimental results. Note that you can also use our proof-of-concept system in http://dslab.snu.ac.kr/demo.html.

• Journal of Service Research and Studies
• /
• v.13 no.2
• /
• pp.41-61
• /
• 2023

The universal form of life in the era of the 4th industrial revolution can probably be summarized as the keyword "non-face-to-face". In particular, in terms of consumption activities, face-to-face contact is gradually changing to a system that minimizes, and offline stores are rapidly changing to non-contact services through kiosks and robots. The social structure is also changing with the passage of time, and most fundamentally, our dietary consumption patterns are changing. In particular, the increase in single-person households and the aging population are having a great impact on changes in the food service industry, which is closely related to dietary life. The HMR (Home Meal Replacement) market has grown significantly as the labor of cooking at home has decreased and the use of substitute foods has increased. As the size of the market has grown, the types of businesses that provide products have also diversified. The development of technology, non-face-to-face culture, and corporate management efficiency are intertwined, and unmanned stores are spreading recently. In this study, service quality attributes of HMR unmanned stores, where competition is gradually intensifying, are classified, and service quality classification using the Kano model and Timko's customer satisfaction coefficient are calculated to provide implications for service management based on customer satisfaction. As a result of the analysis, 'products with short cooking time' and 'variety of products (menu)' were classified as attractive qualities, and 'cleanliness inside/outside of the store' and 'products at reasonable prices' were classified as unified quality. In addition, 'convenience of self-checkout process' was classified as a natural quality, and 'convenience of in-store passage' was classified as an indifferent quality. Furthermore, when the service factor was satisfied within the HMR unmanned store, the factor with the highest satisfaction coefficient was 'product (menu) variety', and the factor with the highest dissatisfaction factor was 'convenience of self-checkout process'. Through the results of this study, it is intended to derive priorities in service quality management of HMR unmanned stores and provide strategic implications for related businesses.