• Knowledge Management Research
• /
• v.22 no.4
• /
• pp.27-40
• /
• 2021

Click-Through Rate (CTR) prediction is a key function that determines the ranking of candidate items in the recommendation system and recommends high-ranking items to reduce customer information overload and achieve profit maximization through sales promotion. The fields of natural language processing and image classification are achieving remarkable growth through the use of deep neural networks. Recently, a transformer model based on an attention mechanism, differentiated from the mainstream models in the fields of natural language processing and image classification, has been proposed to achieve state-of-the-art in this field. In this study, we present a method for improving the performance of a transformer model for CTR prediction. In order to analyze the effect of discrete and categorical CTR data characteristics different from natural language and image data on performance, experiments on embedding regularization and transformer normalization are performed. According to the experimental results, it was confirmed that the prediction performance of the transformer was significantly improved when the L2 generalization was applied in the embedding process for CTR data input processing and when batch normalization was applied instead of layer normalization, which is the default regularization method, to the transformer model.

• Journal of the Korean Society of Marine Environment & Safety
• /
• v.28 no.4
• /
• pp.542-548
• /
• 2022

In order to reduce marine accidents, each contracting Government needs to implement the instruments enacted and amended by the International Maritime Organization (IMO). The III Code requires each administration of the government to have a system for improvement through periodic review and evaluation and to include performance indicators in its evaluation methods. Thus, each IMO Member State needs to develop its own performance indicators. The purpose of this paper is to develop and present an evaluation model using the Balanced Scorecard (BSC) and Key Performance Indicators (KPI) in order to quantify and evaluate the level of implementation of the instruments by the administrations. From the perspective of 'III-BSC', which applies the BSC concept to the III code requirements, the Critical Success Factors (CSF) that must be secured to achieve the established vision were drawn up, and candidate KPIs for each evaluation area were developed to measure the derived key success factors and an initial study model was designed composed of four levels. The validity of the KPIs was verified and the study model was finalized using the survey design using the SMART technique. Furthermore, based on the developed study model, an evaluation model for the implementation of the BSC-based IMO instruments was developed by deriving the weights of elements for each level through AHP analysis. The developed evaluation model is expected to contribute toward improving the administrations' level of implementation of the IMO instruments as a tool for quantitatively grasping the level of performance of the implementation.

• Journal of Intelligence and Information Systems
• /
• v.28 no.2
• /
• pp.263-278
• /
• 2022

With the recent advent of IoT technology, automatic pet feeders are being distributed so that owners can feed their companion animals while they are out. However, due to behaviors of pets, the method of measuring weight, which is important in automatic feeding, can be easily damaged and broken when using the scale. The 3D camera method has disadvantages due to its cost, and the 2D camera method has relatively poor accuracy when compared to 3D camera method. Hence, the purpose of this study is to propose a deep learning approach that can accurately estimate weight while simply using a 2D camera. For this, various convolutional neural networks were used, and among them, the ResNet101-based model showed the best performance: an average absolute error of 3.06 grams and an average absolute ratio error of 3.40%, which could be used commercially in terms of technical and financial viability. The result of this study can be useful for the practitioners to predict the weight of a standardized object such as feed only through an easy 2D image.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.1
• /
• pp.51-58
• /
• 2022

Drivable area detection, one of the main functions of advanced driver assistance systems, means detecting an area where a vehicle can safely drive. The drivable area detection is closely related to the safety of the driver and it requires high accuracy with real-time operation. To satisfy these conditions, V-disparity-based method is widely used to detect a drivable area by calculating the road disparity value in each row of an image. However, the V-disparity-based method can falsely detect a non-road area as a road when the disparity value is not accurate or the disparity value of the object is equal to the disparity value of the road. In a road environment including vegetation, such as a highway and a country road, the vegetation area may be falsely detected as the drivable area because the disparity characteristics of the vegetation are similar to those of the road. Therefore, this paper proposes a drivable area detection method and hardware architecture with a high accuracy in road environments including vegetation areas by reducing the number of false detections caused by V-disparity characteristic. When 289 images provided by KITTI road dataset are used to evaluate the road detection performance of the proposed method, it shows an accuracy of 90.12% and a recall of 97.96%. In addition, when the proposed hardware architecture is implemented on the FPGA platform, it uses 8925 slice registers and 7066 slice LUTs.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.5
• /
• pp.1-9
• /
• 2022

Localization systems can be used with various circumstances like measuring population movement and rescue technology, even in security technology (like infiltration detection system). Vision sensors such as camera often used for localization is susceptible with light and temperature, and can cause invasion of privacy. In this paper, we used ultra-wideband radar technology (which is not limited by aforementioned problems) and machine learning techniques to measure the number and location of occupants in other indoor spaces behind the wall. We used four different algorithms and compared their results, including extremely randomized tree for four different situations; detect the number of occupants in a classroom, split the classroom into 28 locations and check the position of occupant, select one out of the 28 locations, divide it into 16 fine-grained locations, and check the position of occupant, and checking the positions of two occupants (existing in different locations). Overall, four algorithms showed good results and we verified that detecting the number and location of occupants are possible with high accuracy using machine learning. Also we have considered the possibility of service expansion using the oneM2M standard platform and expect to develop more service and products if this technology is used in various fields.

• Journal of Appropriate Technology
• /
• v.7 no.1
• /
• pp.115-124
• /
• 2021

Due to the successful accomplishments of the first-stage base center project, a road-map for the second-stage, global base center 2030 project has recently been proposed. The vision of the base center is to build a technology centered, cooperation based platform for a sustainable global community. The global base center 2030 project is based on three core strategies as well as three key strategies. The main goal of the core strategy is to establish an interdisciplinary smart platform, as well as a global tech-coordination facility to implement sustainable, inclusive, and innovative science and technology based ODA projects. To achieve such goals, the global center will focus on developing a global living lab, interdisciplinary smart linkage systems, and a global operating platform. The main goals for the key strategies are to solve issues at the base centers while establishing an international relationship through sustainable technology. To achieve such goals, key projects are centered in establishing a ICT package, and a global living lab based on smart interconnected system. With this, a global inter-connected business platform will also be established to support technical and operational issues.

• The Korean Journal of Archival Studies
• /
• no.69
• /
• pp.129-152
• /
• 2021

With the enactment and enforcement of the 「Public Records and Archives Act」, the Tottori Prefectural Archives newly enacted the "Records Management Ordinance" through a comprehensive review of the previous archive function. In accordance with this ordinance, which came into effect in April 2012, Tottori Prefecture expanded the records management institutions (Public Security Committee and Police Headquarters, etc.) A series of archive systems were improved, such as the expansion and reinforcement of the authority to appraisal and select. In addition, the Tottori Prefectural Archives went further and implemented the "Ordinance on the Preservation of Historical Documents, etc." from April 2017. Through this, the municipalities unit basic local government's record management support work was set as a unique function of the local archive, and a linkage and cooperation system was established for the preservation of private records of the prefecture area as well as the basic local area together with cultural heritage institutions such as museums and libraries. As a reference case that continuously guarantees the performance of various activities based on the mission and vision of the local archives in Korea that aim for 'autonomy of records' on the poor archival culture soil, it is worth paying attention to the case of continuous record management reform of the Tottori Prefectural Archives through the enactment of the original role and function of the archive.

• Iberoamérica
• /
• v.22 no.2
• /
• pp.25-57
• /
• 2020

Eduardo Galeano is a left-wing intellectual who led the criticism and accusations of dictatorship and social absurdity in Latin America. It digs into the truth of hidden history that has not been revealed in official history. He values the memory of history to stop repetition of the unfortunate history of the past. The main research topic of this study, 『The Book of Embraces』, is also an extension of such work. Most of the stories in this work depend on the writer's memory. There is no coherence or integration in the content of the story, and the length of the text is not constant, so it is extremely informal and fragmented. This is a strategy to formally reveal the illogical and irrational reality of Latin America. He analyzes the problems of the separation system prevalent in Latin American society from various perspectives. It separates me and the others as well as the past and the present. It makes the memory of history void and paralyzes the consciousness of history. These systems are fixed for convenient governance. In this situation, the pattern of violence becomes more explicit and broad. The anxiety and fear of the Latin American public become commonplace. It is a reality of enduring daily life without hope. Galeano finds this enduring force in historical memory. He believes that when the past and the present meet and embrace, a new history of the future can be encountered. Galeano does not just criticize reality or cynical attitude but also suggests hope for the future.

• Journal of Korean Elementary Science Education
• /
• v.41 no.2
• /
• pp.173-185
• /
• 2022

The purpose of this study is to present a reflective review of the earth and universe units from the revised elementary curriculum of 2007-2015 and suggest changes in the 2022 revised curriculum. For this purpose, we conducted an FGI with earth science educators and elementary school teachers regarding the content elements and system, the achievement standards and inquiry activity composition, and the vertical and horizontal curriculum connectivity. Free response and weighted hierarchical analysis items were incorporated into the FGI to ensure logical consistency of the inductively derived improvement. This analysis revealed that the composition of units by grade group had been unevenly distributed among each of the "earth systems" until the 2015 revised curriculum was finalized. Furthermore, the basic concept was still insufficient. We suggest that achievement standards centered on the learning content and skills must state specific scientific core competencies, and inquiry activities should include rigorous critical thinking, student written responses, and student inquiry and analysis. In the hierarchical analysis items, FGI emphasized the inclusion of essential content elements rather than reduction of content elements, understanding-oriented concept learning rather than interest-centered phenomenon learning, basic concept division learning before integration between subjects, and expanding vertical-horizontal connectivity rather than repeating and advancing learning. There is a limit to the generalizing the suggestions proposed in this study to the common opinion of elementary earth science experts. However, since the main vision of the 2022 revised curriculum is to gather opinions through educational entities' participation in a variety of educational subjects, it is suggested that our results should be incorporated as one of the opinions proposed for the 2022 curriculum revision.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.21 no.5
• /
• pp.274-286
• /
• 2022

Multi-object tracking has been studied for a long time under computer vision and plays a critical role in applications such as autonomous driving and driving assistance. Multi-object tracking techniques generally consist of a detector that detects objects and a tracker that tracks the detected objects. Various publicly available datasets allow us to train a detector model without much effort. However, there are relatively few publicly available datasets for training a tracker model, and configuring own tracker datasets takes a long time compared to configuring detector datasets. Hence, the detector is often developed separately with a tracker module. However, the separated tracker should be adjusted whenever the former detector model is changed. This study proposes a system that can train a model that performs detection and tracking simultaneously using only the detector training datasets. In particular, a Siam network with augmentation is used to compose the detector and tracker. Experiments are conducted on public datasets to verify that the proposed algorithm can formulate a real-time multi-object tracker comparable to the state-of-the-art tracker models.