Vol. 47, No. 12,
Dec. 2020
Digital Library
Case Studies and Trends in Data Reproduction and Distributed Processing using Event Sourcing and CQRS Pattern
Sangkon Han, Jung-in Choi, Gyun Woo
http://doi.org/10.5626/JOK.2020.47.12.1101
The amounts of information collected in areas such as V2X, artificial intelligence, and self-driving are increasing rapidly every year. It is important to collect, process, and store heterogeneous data. This is because the amounts of relevant data have increased rapidly and the types and formats of those data have diversified. Event sourcing, which is a way to store a whole set of events about changing the state of a system or an application through some set of events, enables an event-driven design based on messages, and also enables the restoration and reproduction of the state of a system or an application. Combined with CQRS (Command and Query Responsibility Segmentation), a pattern that separates queries and commands, the characteristics of event sourcing can be utilized for a distributed system architecture for large data processing, as well as systems and applications for data analysis and debugging using reproducibility of the state of event sourcing. This paper introduces the cases and research contents on data reproduction and dispersal processing using event sourcing.
Dynamic Core Affinity for Manycore Partitioning
Chan-Gyu Lee, Joong-Yeon Cho, Hyun-Wook Jin
http://doi.org/10.5626/JOK.2020.47.12.1111
As the number of cores in computer systems with NUMA architecture is increasing, the contemporary operating systems are not scalable because of increased cache misses, cache coherence activities and synchronizations. To resolve this problem, several studies have suggested controlling the core affinity of the system calls and the event handlers, making these run on a specific set of cores. However, these core partitioning approaches statically decide the number of cores available for controlling the core affinity without considering the characteristics of the applications and the system architectures. In this paper, we propose a dynamic core affinity scheme for the core partitioning and compare with a static core partitioning mechanism.
Prediction of Blood Glucose in Diabetic Inpatients Using LSTM Neural Network
Sang Hyeon Kim, Han Beom Lee, Seong Wan Jeon, Dae Yeon Kim, Sang Jeong Lee
http://doi.org/10.5626/JOK.2020.47.12.1120
Diabetes is a chronic disease that causes serious complications, and at the medical site, doctors predict future changes in blood glucose based on patients past blood glucose trends and implement medical treatment. Recently, a CGM(Continuous Glucose Monitoring) measuring device has been introduced that can automatically measure blood glucose every five minutes to monitor continuous changes in blood glucose, and it is widely used in clinical applications. Based on the results of CGM blood glucose, the doctors predict and treat the timing of insulin administration and high risk of diabetes patients. In this paper, the blood glucose prediction model based on deep learning neural network is proposed. The proposed model is designed with an LSTM (Long Short-Term Memory) based neural network. It is designed to take historical blood glucose data as well as variables such as HbA1c(glycated hemoglobin) and BMI(body mass index). It was applied and tested using CGM blood glucose data from Type 2 Diabetes inpatients at a university hospital. The proposed model which patient characteristics show50% improvement at maximum in blood glucose prediction accuracy over the LSTM model of previous study.
Long-distant Coreference Resolution by Clustering-extended BERT for Korean and English Document
Cheolhun Heo, Kuntae Kim, Key-sun Choi
http://doi.org/10.5626/JOK.2020.47.12.1126
Coreference resolution is a natural language processing task of identifying all mentions that refer to the same denotation in the given natural language document. It contributes to improving the performance of various natural language processing tasks by resolving the co-referents caused by linguistically replaceable realizations by using the referencible forms such as pronouns, indicative adjectives, and abbreviations but preventing co-referencing of homonyms (i.e., same form but different meaning). We propose a novel approach to coreference resolution particulary to identify the long-distant co-referents by applying long-distance clustering for surface forms under a BERT-based model performing well in English. We compare the performance of the proposed model and other models over the Korean and English datasets. Results demonstrated that our model has a better grasp of contextual elements compared to the other models.
Korean Abstract Meaning Representation (AMR) Guidelines and Corpus for Graph-structured Meaning Representations
Hyonsu Choe, Jiyoon Han, Hyejin Park, Taehwan Oh, Seokwon Park, Hansaem Kim
http://doi.org/10.5626/JOK.2020.47.12.1134
This paper introduces the Korean Abstract Meaning Representation (AMR) Guideline v1.0. AMR is a graph-based meaning representation system and is one of the most significant frameworks for meaning representation. The Korean AMR Guideline is a product of the study that analyzed and localized the AMR Guideline 1.2.6 on the basis of the features of the Korean language has. The Korean AMR corpus can be used for implementation of semantic parser, which is the core of Natural Language Understanding technology, and can be used for NLU/NLG tasks such as Machine Reading Comprehension, Automatic Summarization. The Korean AMR Corpus built depending on this guideline comprises 896 sentences, or 10,414 words (eojeol) for now.
A Study on Development of VR Content Creation System for Improving Conjugal Relation
http://doi.org/10.5626/JOK.2020.47.12.1142
Currently, the divorce rate in Korean society is increasing continuously. The problem is that there are few remedies to improve marital relations, and most remedies rely on professional counselors. One of the counseling methods used by professional counselors to improve the relationship is to enable improvement of the relationship by reminding partners of past times. Among the IT technologies that can enable recalling a certain point in the past, is using VR technology. This paper discusses a system that can generate VR content for improving conjugal relations. Based on the memories of the target couples, an editor for generating VR content based on a timeline was developed and a module for assessing if a relationship improved when VR content developed as an editor of a paper were used was developed and implemented as a system. The VR content generation system in this paper is expected to be used to improve marital relationships, as well as maximize productivity and convenience when creating various VR content based on user memories, which will be used in various applications in other fields.
An Embedding Method of Emotes for the Detection of Popular Clips on Twitch.tv
Hyeonho Song, Kunwoo Park, Meeyoung Cha
http://doi.org/10.5626/JOK.2020.47.12.1153
This study presents an embedding method that effectively learns emote’s meaning in Twitch.tv to understand the audience reaction in live streaming. The proposed method first trains an embedding matrix for text and emotes, respectively, and merges the two matrices into one. Using 2,220,761 clips shared on Twitch.tv, this study conducted two experiments: clustering and clip popularity prediction. Results showed that the approach identifies emote clusters that express a similar emotion and detects popular clips. Future studies could utilize the proposed emote embedding method for the highlight prediction of a live stream.
An Integrated System for Large-Scale Knowledge Graph Inference Using the Spark DataFrame
Min-Ho Lee, Min-Sung Kim, Wan-Gon Lee, Young-Tack Park
http://doi.org/10.5626/JOK.2020.47.12.1162
Recently, there has been an active study of large-scale ontology reasoning methods using big data obtained from the Web. However, when the amount of data increases, there is a problem with inference performance and processing speed decreasing. In this paper, we propose a two-step integrated system to perform inference using the Spark DataFrame in a cloud computing environment for effective inference. The first step is to perform rule inference on the OWL through a previous study inference engine. The second step, as in the previous study, performs inference on the user-defined rules through the SWRL inference engine using the Spark DataFrame.
Query-based Abstractive Summarization Model Using Sentence Ranking Scores and Graph Techniques
http://doi.org/10.5626/JOK.2020.47.12.1172
The purpose of the fundamental abstractive summarization model is to generate a short summary document that includes all important contents within the document. Conversely, in the query-based abstractive summarization model, information related to the query should be selected and summarized within the document. The existing query-based summarization models calculates the importance of sentences using only the weight of words through an attention mechanism between words in the document and the query. This method has a disadvantage in that it is difficult to reflect the entire context information of the document to generate an abstractive summary. In this paper, we resolve this problems by calculating the sentence ranking scores and a sentence-level graph structure. Our proposed model shows higher performance than the previous research model, 1.44%p in ROUGE-1 and 0.52%p in ROUGE-L.
Survey on Feature Attribution Methods in Explainable AI
Gihyuk Ko, Gyumin Lim, Homook Cho
http://doi.org/10.5626/JOK.2020.47.12.1181
As artificial intelligence (AI)-based technologies are increasingly being used in areas that can have big socioeconomic effects, there is a growing effort to explain decisions made by AI models. One important direction in such eXplainable AI (XAI) is the ‘feature attribution’ method, which explains AI models by assigning a contribution score to each input feature. In this work, we surveyed nine recently developed feature attribution methods and categorized them using four different criteria. Based on the categorizations, we found that the current methods focused only on specific settings such as generating local, white-box explanations of neural networks and lacked theoretical foundations such as axiomatic definitions. We suggest future research directions toward a unified feature attribution method based on our findings.
A Method for Training Data Selection based on LSTRf
Myunggwon Hwang, Yuna Jeong, Wonkyung Sung
http://doi.org/10.5626/JOK.2020.47.12.1192
This paper presents a data selection method that has a positive effect on learning for an efficient human-in-the-loop (HITL) process required for automated and intelligent artificial intelligence (AI) development. Our method first maps the training data onto a 2D distribution based on similarity, and then grids are laid out with a fixed ratio. By applying Least Slack Time Rate first (LSTRf) techniques, the data are selected based on the distribution consistency of the same class data within each grid. The finally selected data are used as convolutional neural network (CNN)-based classifiers to evaluate the performance. We carried out experiments on the CIFAR-10 dataset, and evaluated the effect of grid size and the number of data selected in one operation. The selected training data were compared to randomly selected data of the same size. The results verified that the smaller the grid size (0.008 and 0.005) and the greater the number selected in the single operation, the better the learning performance.
A Trie-based Indexing Scheme for Efficient Retrieval of Massive Spatio-Temporal IoT Sensor Data
Hawon Chu, Young-Kyoon Suh, Ryong Lee, Minwoo Park, Rae-Young Jang, Sang-Hwan Lee, Sa-Kwang Song
http://doi.org/10.5626/JOK.2020.47.12.1199
As the Internet-of-Things (IoT) sensors with enhanced communication technology and computing power have been widely utilized in many areas, a great deal of spatio-temporal data has been continuously generated. Thanks to the remarkable advances in storage technology, it is possible to collect such massive data into storage systems for further high-dimensional analysis. That said, it has been very challenging to speedily locate stored IoT data in a reasonable amount of time due to the heavy volume and complex spatial and temporal attributes. To address this concern, we propose a novel scalable indexing scheme, termed ST-Trie, to support the efficient querying of massive spatial-temporal data collected from IoT sensors. The key idea of our scheme is to encode three-dimensional spatiotemporal information into one-dimensional keys in consideration of time and space locality and then organize the keys into a logical trie structure. In our experiments with real datasets, the proposed scheme outperformed composite indexes by an average of up to 92 times in terms of query response time. In particular, we confirmed that ST-Trie scaled much better than the compared indexes with increasing time ranges.
A Study on the Intelligent Delivery Management System Using UAV-Edge Computing Technology
Chu Myaet Thwal, Minkyung Lee, Choong Seon Hong
http://doi.org/10.5626/JOK.2020.47.12.1208
With the recent advancement of the digital and Internet-of-Things (IoT) technologies, cities globally are rapidly transforming into smart cities. In a parallel to the IoT technology, another technology that has substantially improved in recent years is the Unmanned Aerial Vehicle (UAV) technology, resulting in cheaper, more powerful and reliable UAVs. In this paper, we integrate the aid of the IoT and propose an intelligent delivery management system in coordination with edge computing and UAV technology. The proposed system is an innovative system that facilitates in reducing the operational delay of the delivery services and provides greater and faster facilities to customers. The whole procedure of the delivery process is managed by the edge-based control stations serving as the media between the retailers and UAVs. These stations are distributed across urban areas and are responsible for assigning tasks to the UAVs by performing crucial calculations and communications between the retailers and UAVs. By applying the proposed intelligent delivery scheme in smart city applications, it can be expected to reduce delays in delivery services because of the shortage of manual labor and traffic conditions, thus providing greater and faster facilities to the customers.
Search
Journal of KIISE
- ISSN : 2383-630X(Print)
- ISSN : 2383-6296(Electronic)
- KCI Accredited Journal
Editorial Office
- Tel. +82-2-588-9240
- Fax. +82-2-521-1352
- E-mail. chwoo@kiise.or.kr