Jo Jul 31, 2026
The displacement ventilation system (DVS) is one of the mechanical ventilation systems which blow the air with lower temperature than indoor air from the lower part of a room and cause upward moving of indoor air by buoyant force to exhaust it at the ceiling.
As DVS is aimed at meeting only the requirements of workplaces, the thermal and concentration stratification is formed between lower and upper parts. DVS makes the quality of the air better and exhausts the polluted air effectively, and therefore the saving efficiency of energy is remarkable. In the designing of DVS, the height of thermodynamic stratification has to be controlled to be higher than that of the workplace for people, but it should not be too high. This is because being too high might cause an increment in the amount of air blast, which results in waste. Thus, the determination of the height of thermodynamic stratification is one of the vital problems in the application of DVS.
Paek Myong Chol, a section head at the Faculty of Thermal Engineering, analyzed the characteristics of indoor airflow and temperature distribution in the hybrid displacement ventilation system (HDVS) where a vertical supply duct was installed at the middle height of a building by computational fluid dynamics (CFD). Then, he investigated the specific influence of the velocity of air inflow and the length of the supply duct on the air temperature.
The results show that the proposed system saves 35kw of energy compared to the mixing ventilation system, which leads to the annual energy saving of about 80 000kW·h.
If more information is needed, you can refer to his paper “Energy Saving Characteristics of Hybrid Displacement Ventilation System with Vertical Supply Duct” in “GAS Journal of Engineering and Technology” (SCI).
...
Jo Jul 30, 2026
All the digital data stored, copied and distributed on the network have their own specific information, and digital images, as one of the fastest multimedia formats that can deliver their information in a visual way, particularly have a great amount of information. Most of the digital images transmitted through network are private, and especially, those with sensitive information such as medical diagnostic images can lead to irrevocable disaster when they are disclosed to be abused. Therefore, it is very important for digital images to be secured.
In general, digital images, when compared to text data, are characterized by very large capacity, high overlap, and strong correlation between adjacent pixels. On the other hand, digital image data requires strong real-time property in communication, storage and distribution, and hence, it requires a prompt and secure image encryption algorithm. From this necessity, many precedent researchers have proposed some improved chaotic maps and applied them to various image encryption algorithms.
Song Ok Chol, an institute head at the Institute of Information Technology, configured a novel 3D-ICCM (3D Infinite Collapse Coupling Map) with better chaotic performance by combining three 1D infinite collapse maps that has the best chaotic characteristics among the existing 1D chaotic maps, and analyzed its chaotic characteristics through a comparison with the previous 3D-ICM. Then, he proposed a new simple-structured pixel-level image encryption algorithm and evaluated its performance by applying the proposed 3D-ICCM.
The experimental simulation and analysis show that the proposed 3D-ICCM and image encryption algorithms have excellent performance.
For more information, please refer to his paper “Robust pixel-level image encryption algorithm using 3D infinite collapse coupling map” in “Multimedia Tools and Applications” (SCI).
...
Jo Jul 29, 2026
In the past, users only found the information they needed on the Web, but recently newly-developed information services give convenience to users by automatically providing useful information for users thanks to the development of the Web.
Recently, with the development of cloud computing platforms, mobile Internet applications are constantly emerging. Mobile retrieval is one of the most important applications of mobile Internet. Mobile search must be personalized. This is because different users need different information depending on their environments and interests even if the query required is the same.
Recommendation systems are widely used on the Web as this kind of service. The recommendation system is a personalized information filtering system that provides users with new items that might please users based on the assessment information or purchase information about items that users have already experienced.
Recently, recommendation systems have been adopted in various domains such as recommendation of commodities that users wants to buy, recommendation of films, TV, broadcast, music, restaurants, etc., identification of web pages of interest to users, and selection of appropriate information retrieval methods. However, insufficient user’s personal information and frequent use of other computers increase the difficulty in collecting users’ personal information. Due to the nature of mobile phones in mobile environments, user’s personal information is easily collected by telecommunication and it is no longer a difficult problem.
With a view to further increasing the effectiveness of personalized mobile retrieval method, Kim Kwang Hyok, a researcher at the Institute of Information Technology, proposed a hybrid filtering recommendation method by combining content-based filtering and cooperative filtering.
The experiments demonstrate that the proposed method can significantly improve the effectiveness of personalized mobile retrieval.
You can find his paper “Study on Personalization Information Retrieval Using Hybrid Filtering Recommendation Method” in “2025 10th International Conference on Intelligent Information Technology (ICIIT 2025)”.
...
Jo Jul 28, 2026
Chaotic map which has excellent unpredictability and complexity is extensively used in various fields such as information security, information communication, system control, etc., and thus, there is an intensive research work to improve its chaotic characteristics.
Generally, digital images, compared to text data, have some characteristics different from ordinary data, such as very large capacity, high overlap and strong correlation between adjacent pixels. On the other hand, digital image data requires strong real-time property in communication, storage and distribution, and hence a fast and secure algorithm is needed for its encryption.
Jong Chol Min, a researcher at the Institute of Information Technology, studied the excellent chaotic performance of 2D-ICCM (2D Infinite Collapse Coupling Map) proposed by a researcher and proposed a simple-structured and fast-speed pixel-level image encryption algorithm using 2D-ICCM.
In image encryption algorithms using chaotic maps, it is common to repeat permutation and diffusion processes several times in order to increase the complexity of encryption and decryption processes. However, it results in degradation of the whole performance of encryption scheme, especially degradation of the ability to resist against cropping and noise attacks in the decryption process. Therefore, he bolstered the system complexity by using the chaotic output sequence used in each step interchangeably while repeating the permutation and diffusion process only twice.
He verified through various experiments that the pixel-level image encryption algorithm using 2D-ICCM has comparatively better chaotic performance than the previous 2D chaotic maps and algorithms.
His paper “A New Color Image Encryption Algorithm Using 2D Chaotic Map” was presented at “2025 10th International Conference on Intelligent Information Technology (ICIIT 2025)”.
...
Jo Jul 27, 2026
In neural machine translation (NMT), BPE is usually used to limit vocabularies since a large number of vocabularies have a negative impact on training. Use of BPE could limit the number of vocabularies to some extent, but word alignment by attention mechanism cannot be done correctly since vocabularies are not properly segmented into word stems.
In neural machine translation, a huge number of vocabularies from the corpus have a negative impact on its training, and the longer the length of sentences which reflects the number of tokens is, the longer the training time gets. Therefore, it is important to perform good tokenization for minimizing the number of vocabularies and the length of sentences.
Kim Tong Gwon, a researcher at the Institute of Information Technology, proposed a stem and ending based tokenization (SEBT) method for improving the accuracy of word alignment by applying to tokenization a stem and ending dictionary which was made by segmenting BPE tokens into stems and endings.
This method has four advantages:
First, it is possible to reduce the number of vocabularies by correctly segmenting the stems and endings of Russian and Korean.
Second, it can improve learning efficiency by relieving the sparse word problem.
Third, it can improve the alignment accuracy between Russian and Korean tokens in the attention mechanism.
Fourth, it can reduce the training time since the sentence length is not increased after tokenization.
The experiments showed that this approach improves the performance of Russian-Korean NMT by improving the alignment between Russian and Korean tokens during the training of Russian-Korean neural machine translation and by reducing the number of Russian and Korean vocabularies.
You can find more information in his paper “SEBT: Word Stem and Ending based Tokenization in Russian-Korean Neural Machine Translation” in “2025 10th International Conference on Intelligent Information Technology (ICIIT 2025)”.
...
Jo Jul 26, 2026
Since the impact performance of a drill is the main parameter for determining the drilling efficiency, study of the impact performance of a drill has a crucial influence on its productivity. As the impact energy of a rock drill is an important characteristic to evaluate its performance, many researchers have proposed various methods of measuring impact energy in order to evaluate it accurately.
Evaluating the impact performance of a rock drill by their methods requires a standardized rock sample in advance. Therefore, it is difficult to apply them to the performance test of a rock drill in batch production processes.
Yang Un Hyok, a section head at the Faculty of Mining Engineering, proposed a new structure of measuring device for estimating the impact energy of a percussive drill and analyzed its characteristics.
In the newly proposed measurement device, the maximum value of the pressure wave generated by the impact of 33-56J is approximately linear, with a range of 12-15MPa. It is shown that the error in the new proposed measurement device is less than 5% when the oil temperature is 10-70℃, air content is less than 2% and oil leakage is 10%.
The results of the study show that the proposed device ensures high measurement accuracy and significantly reduces damage to the sensor.
For more information, please refer to his paper “Influence of Oil Temperature, Air Content and Leakage in Hydraulic Measuring Device for Estimating Impact Energy of Rock Drill” in “Proceedings of KUTIC-2025”.
...