• Journal of Korean Society for Geospatial Information Science
• /
• v.24 no.1
• /
• pp.25-33
• /
• 2016

In order to visualize point based Location-Based Social Network Services(LBSNS) data on multi-scaled tile map effectively, it is necessary to apply tile-based clustering method. Then determinating reasonable numbers and size of tiles is required. However, there is no such criteria and the numbers and size of tiles are modified based on data type and the purpose of analysis. In other words, researchers' subjectivity is always involved in this type of study. This is when Modifiable Areal Unit Problem(MAUP) occurs, that affects the results of analysis. Among LBSNS, geotagged Twitter data were chosen to find the influence of MAUP in scale effects perspective. For this purpose, the degree of spatial autocorrelation using spatial error model was altered, and change of distributions was analyzed using Morna's I. As a result, positive spatial autocorrelation showed in the original data and the spatial autocorrelation was decreased as the value of spatial autoregressive coefficient was increasing. Therefore, the intensity of the spatial autocorrelation of Twitter data was adjusted to five levels, and for each level, nine different size of grid was created. For each level and different grid sizes, Moran's I was calculated. It was found that the spatial autocorrelation was increased when the aggregation level was being increased and decreased in a certainpoint. Another tendency was found that the scale effect of MAUP was decreased when the spatial autocorrelation was high.

• Journal of Korean Society of Transportation
• /
• v.23 no.1
• /
• pp.103-114
• /
• 2005

The transit fare resulted from the renovation of public transit system in Seoul is basically determined based on the distance-based fare policy (DFP). In DFP, the total fare a passenger has to pay for is calculated by the basic-transfer-premium fare decision rule. The fixed amount of the basic fare is first imposed when a passenger get on a mode and it lasts within the basic travel distance. The transfer fare is additionally imposed when a passenger switches from one mode to another and the fare of the latter mode is higher than the former. The premium fare is also another and the fare of the latter begins to exceed the basic travel distance and increases at the proportion of the premium fare distance. The purpose of this study is to propose an algorithm for finding K number of paths, paths that are sequentially sorted based on total amount of transit fare, under DFP of the idstance-based fare policy. For this purpose, the link mode expansion technique is proposed in order to save notations associated with the travel modes. Thus the existing K shortest path algorithms adaptable for uni-modal network analysis are applicable to the analysis for inter-modal transportation networks. An optimality condition for finding the K shortest fare routes is derived and a corresponding algorithms is developed. The case studies demonstrate that the proposed algorithm may play an important role to provide diverse public transit information considering fare, travel distance, travel time, and number of transfer.

• Journal of KIISE:Databases
• /
• v.33 no.1
• /
• pp.138-149
• /
• 2006

As the environment with sensor network and ubiquitous computing is emerged, there are many demands of handling continuous, fast data such as streaming data. As work about streaming data has begun, work about management of streaming data in Publish-Subscribe system is started. The recent emergence of XML as a standard for information exchange on Internet has led to more interest in Publish - Subscribe system. A filtering technique of streaming XML data in the existing Publish- Subscribe system is using some schemes based on automata and YFilter, which is one of filtering techniques, is very popular. YFilter exploits commonality among path queries by sharing the common prefixes of the paths so that they are processed at most one and that is using the top-down approach. However, because partial matching path queries interrupt the common prefix sharing and don't calculate from root, throughput of YFilter decreases. So we use sharing of commonality among path queries with the common postfixes of the paths and use the bottom-up approach instead of the top-down approach. This filtering technique is called as PoSFilter. And we verify this technique through comparing with YFilter about throughput.

• Journal of the Korean Regional Science Association
• /
• v.37 no.3
• /
• pp.3-18
• /
• 2021

The fourth Industrial Revolution is transforming the industrial structure of the region, and it is necessary to develop new industries and technologies that reflect regional characteristics. The purpose of this study is to measure the knowledge relatedness and technological complexity in Busan, Ulsan, and Gyeongnam, and to identify technologies with potential for regional industrial differentiation strategies. Using patent data from 2015 to 2019, co-occurrence matrices were derived from 652 IPC codes, and the knowledge relatedness density and technology complexity index were calculated. Network analysis was performed using the knowledge relatedness density. As a result of analysis, it was found that mechanical engineering occupied a large proportion, followed by chemistry and electrical engineering. As a result of applying the risk-benefit framework to derive technologies with the potential to differentiate local industries, the technological capabilities of low-risk-high-benefit were different. Among mechanical engineering, technologies such as engine, machine operation, and transportation were included in Busan. In Ulsan, environmental technology in chemical and materials, and heat treatment technology in mechanical engineering were technologies with low-risk and high-benefit capabilities. Gyeongnam showed competence in mechanical engineering, chemistry, and electrical engineering in some areas such as Gimhae, Yangsan, and Changwon. The results of this study are meaningful in that they identified technologies with potential for selecting and deriving strategic industries for regional growth based on latent knowledge in the region.

• Journal of Broadcast Engineering
• /
• v.26 no.2
• /
• pp.125-131
• /
• 2021

In recent years, Convolutional Neural Networks (CNNs) have achieved outstanding performance in the fields of computer vision such as image classification, object detection, visual quality enhancement, etc. However, as huge amount of computation and memory are required in CNN models, there is a limitation in the application of CNN to low-power environments such as mobile or IoT devices. Therefore, the need for neural network compression to reduce the model size while keeping the task performance as much as possible has been emerging. In this paper, we propose a method to compress CNN models by combining matrix decomposition methods of LR (Low-Rank) approximation and CP (Canonical Polyadic) decomposition. Unlike conventional methods that apply one matrix decomposition method to CNN models, we selectively apply two decomposition methods depending on the layer types of CNN to enhance the compression performance. To evaluate the performance of the proposed method, we use the models for image classification such as VGG-16, RestNet50 and MobileNetV2 models. The experimental results show that the proposed method gives improved classification performance at the same range of 1.5 to 12.1 times compression ratio than the existing method that applies only the LR approximation.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.29 no.2
• /
• pp.377-391
• /
• 2019

Recently, system studies using tokens and block chains for authentication, access control, etc in IoT environment have been going on at home and abroad. However, existing token-based systems are not suitable for IoT environments in terms of security, reliability, and scalability because they have centralized characteristics. In addition, the system using the block chain has to overload the IoT device because it has to repeatedly perform the calculation of the hash et to hold the block chain and store all the blocks. In this paper, we intend to manage the access rights through tokens for proper access control in the IoT. In addition, we apply the Tangle to configure the P2P distributed ledger network environment to solve the problem of the centralized structure and to manage the token. The authentication process and the access right grant process are performed to issue a token and share a transaction for issuing the token so that all the nodes can verify the validity of the token. And we intent to reduce the access control process by reducing the repeated authentication process and the access authorization process by reusing the already issued token.

• Journal of Digital Convergence
• /
• v.17 no.3
• /
• pp.107-117
• /
• 2019

In the era of SNS, many people relies on it to express their emotions about various kinds of products and services. Therefore, for the companies eagerly seeking to investigate how their products and services are perceived in the market, emotion mining tasks using dataset from SNSs become important much more than ever. Basically, emotion mining is a branch of sentiment analysis which is based on BOW (bag-of-words) and TF-IDF. However, there are few studies on the emotion mining which adopt feature selection (FS) methods to look for optimal set of features ensuring better results. In this sense, this study aims to propose FS methods to conduct emotion mining tasks more effectively with better outcomes. This study uses Twitter and SemEval2007 dataset for the sake of emotion mining experiments. We applied three FS methods such as CFS (Correlation based FS), IG (Information Gain), and ReliefF. Emotion mining results were obtained from applying the selected features to nine classifiers. When applying DT (decision tree) to Tweet dataset, accuracy increases with CFS, IG, and ReliefF methods. When applying LR (logistic regression) to SemEval2007 dataset, accuracy increases with ReliefF method.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.1
• /
• pp.45-55
• /
• 2021

This paper has proposed an algorithm that detecting for dense small vehicle in large image efficiently. It is consisted of two Ensemble Deep-Learning Network algorithms based on Coarse to Fine method. The system can detect vehicle exactly on selected sub image. In the Coarse step, it can make Voting Space using the result of various Deep-Learning Network individually. To select sub-region, it makes Voting Map by to combine each Voting Space. In the Fine step, the sub-region selected in the Coarse step is transferred to final Deep-Learning Network. The sub-region can be defined by using dynamic windows. In this paper, pre-defined mapping table has used to define dynamic windows for perspective road image. Identity judgment of vehicle moving on each sub-region is determined by closest center point of bottom of the detected vehicle's box information. And it is tracked by vehicle's box information on the continuous images. The proposed algorithm has evaluated for performance of detection and cost in real time using day and night images captured by CCTV on the road.

• Journal of the Korea Institute of Information Security & Cryptology
• /
• v.20 no.4
• /
• pp.41-52
• /
• 2010

Message authentication is an essential security element in vehicular ad-hoc network(VANET). For a secure message authentication, integrity, availability, privacy preserving skill, and also efficiency in various environment should be provided. RAISE scheme has been proposed to provide efficient message authentication in the environment crowded with lots of vehicles and generally considered to be hard to provide efficiency. However, as the number of vehicles communicating in the area increases, the overhead is also incurred in proportion to the number of vehicles so that it still needs to be reduced, and the scheme is vulnerable to some attacks. In this paper, to make up for the vulnerabilities in dense vehicular communication network, we propose a more secure and efficient scheme using a process that RSU(Road Side Unit) transmits the messages of neighbor vehicles at once with Bloom Filter, and timestamp to protect against replay attack. Moreover, by adding a handover function to the scheme, we simplify the authentication process as omitting the unnecessary key-exchange process when a vehicle moves to other area. And we confirm the safety and efficiency of the scheme by simulating the false positive probability and calculating the traffic.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.40 no.5
• /
• pp.449-456
• /
• 2022

This study proposes a deep learning algorithm to predict crop yield using GIS (Geographic Information System) to extract soil properties from Soilgrids and soil suitability class maps. The proposed model modified the structure of a published CNN-RNN (Convolutional Neural Network-Recurrent Neural Network) based crop yield prediction model suitable for the domestic crop environment. The existing model has two characteristics. The first is that it replaces the original yield with the average yield of the year, and the second is that it trains the data of the predicted year. The new model uses the original field value to ensure accuracy, and the network structure has been improved so that it can train only with data prior to the year to be predicted. The proposed model predicted the yield per unit area of autumn cabbage for kimchi by region based on weather, soil, soil suitability classes, and yield data from 1980 to 2020. As a result of computing and predicting data for each of the four years from 2018 to 2021, the error amount for the test data set was about 10%, enabling accurate yield prediction, especially in regions with a large proportion of total yield. In addition, both the proposed model and the existing model show that the error gradually decreases as the number of years of training data increases, resulting in improved general-purpose performance as the number of training data increases.