Vol. 48, No. 6,
Jun. 2021
Digital Library
ILP-based Schedule Synthesis of Time-Sensitive Networking
Jin Hyun Kim, Hyonyoung Choi, Kyong Hoon Kim, Insup Lee, Se-Hoon Kim
http://doi.org/10.5626/JOK.2021.48.6.595
IEEE 802.1Qbv Time Sensitive Network (TSN), the latest real-time Ethernet standard, is a network designed to guarantee the temporal accuracy of streams. TSN is an Ethernet-based network system that is actively being developed for the factory automation and automobile network systems. TSN controls the flow of data streams based on schedules generated statically off-line to satisfy end-to-end delay or jitter requirements. However, the generation of TSN schedules is an NP-hard problem; because of this, constraint solving techniques, such as SMT (Satisfiability Modulo Theory) and ILP (Integer Linear Programming), have mainly been proposed as solutions to this problem. This paper presents a new approach using a heuristic greedy and incremental algorithm working with ILP to decrease the complexity of computing schedules and improve the schedule generation performance in computing TSN schedules. Finally, we compare our proposed method with the existing SMT solver approach to show the performance of our approach.
Performance Analysis of Concurrent Multitasking for Efficient Resource Utilization of GPUs
Sejin Kim, Qichen Chen, HeonYoung Yeom, Yoonhee Kim
http://doi.org/10.5626/JOK.2021.48.6.604
As Graphics Processing Units (GPUs) are widely utilized to accelerate compute-intensive applications, their application has expanded especially in data centers and clouds. However, the existing resource sharing methods within GPU are limited and cannot efficiently handle several requests of concurrent cloud users’ executions on GPU while effectively utilizing the available system resources. In addition, it is challenging to effectively partition resources within GPU without understanding and assimilating application execution patterns. This paper proposes an execution pattern-based application classification method and analyzes run-time characteristics: why the performance of an application is saturated at a point regardless of the allocated resources. In addition, we analyze the multitasking performance of the co-allocated applications using smCompactor, a thread block-based scheduling framework. We identify near-best co-allocated application sets, which effectively utilize the available system resources. Based on our results, there was a performance improvement of approximately 28% compared to NVIDIA MPS.
NAFS : Stackable Filesystem for Maximizing Local Access in a NUMA System
Seungjun Ha, Hobin Woo, Euiseong Seo, Beomseok Nam
http://doi.org/10.5626/JOK.2021.48.6.612
Intel Optane DC Persistent Memory has read/write latencies comparable to DRAM but ensures data persistence as in block devices such as SSD. However, Optane DC PM modules are installed in DIMM slots of NUMA nodes but legacy block devices are installed in PCIe or SATA. Therefore, Optane DC PMs are known to suffer from the NUMA effects, and the performance of a multithreaded application depends on the NUMA locality. In this paper, we propose a novel stackable file system, NUMA-Aware Filesystem (NAFS). NAFS divides a file into segment units such that I/Os can be performed in local NUMA nodes where each application thread runs. To enable this feature, NAFS duplicates the file metadata across all NUMA nodes if the number of remote I/Os exceeds a certain threshold. Our performance study shows NAFS reduces the number of accesses to remote NUMA nodes significantly, improving the performance of multithreaded applications.
A Reinforcement Learning-Based Cache Partitioning Scheme for Multi-Core Environments
Donggyu Choi, Jongseok Kim, Honguk Woo, Euiseong Seo
http://doi.org/10.5626/JOK.2021.48.6.618
Most processors currently in use provide the shared last-level cache (LLC). Therefore, when multiple applications compete intensely for the LLC, its hit ratio is adversely impacted by extremely frequent cache line replacement; this may result in a significant degradation of the overall performance. The hardware-based cache partitioning techniques can relieve this issue by isolating the cache space of a core from others. However, it is necessary to use an adaptive and intelligent cache partitioning algorithm to dynamically determine the optimal cache partition. Reinforcement learning is an appropriate approach for this kind of problems. However, its model complexity skyrockets as the number of the applications to partition increases. This paper proposes a reinforcement learning-based cache partitioning scheme that can support a large number of running applications. Firstly, we built a reinforcement learning model and made it learn to perform cache partitioning for a small number of applications. We then extended it by clustering applications with the information obtained via supervised learning models for cache-use characteristic predictions, which enabled cache partitioning for more applications and resulted in performance gains of up to 19.75%.
Program Synthesis for JavaScript via Divide-and-Conquer and Abstraction Interpretation
Jungmin Jo, Hangyeol Cho, Woosuk Lee
http://doi.org/10.5626/JOK.2021.48.6.629
Program synthesis aims to automatically generate a program that satisfies the user intent expressed in the form of a high-level specification. Recent years have witnessed a surge in interest in applying this technology to a wide range of problems. Program synthesis can help improve software development productivity. In this paper, we present an algorithm for synthesizing Javascript programs from input-output examples. Our approach was based on a synergistic combination of the version space algebra-based approach, which can efficiently solve synthesis problems through divide-and-conquer, and abstract interpretation, which can be used to finitize infinite search spaces. We have implemented our approach and evaluated it in 140 problems of synthesizing string- and integer-manipulating programs. On average, the desirable programs were generated within nine seconds.
A Cross-Texting Prevention System Using Syntactic Characteristics of Chat Messages
http://doi.org/10.5626/JOK.2021.48.6.639
Cross-texting refers to accidentally sending a message to an unintended person. It occurs frequently when users chat with multiple counterparts at the same time. Messengers mainly provide a function of canceling sending, but it is only post solution, and users find it difficult to prevent mistakes in advance. In this paper, we proposed a cross-texting detection model by analyzing the syntactic characteristics of chat sentences. It modelizes the previous chat messages of a specific user by extracting the honorifics and completeness features from chat messages, and detects the cross-texting cases by determining whether the target sentences are in accordance with the user chat message model. This approach is significant as it solves the cross-texting detection problem only with syntactic characteristics without semantic analysis by modeling the consistency of the user"s chat attitude. The proposed model detect cross-texting cases with an accuracy of 85.5% from automatically generated data using a real messenger dialogue corpus.
Knowledge Completion System through Learning the Relationship between Query and Knowledge Graph
Min-Sung Kim, Min-Ho Lee, Wan-Gon Lee, Young-Tack Park
http://doi.org/10.5626/JOK.2021.48.6.649
The knowledge graph is a network comprising of relationships between the entities. In a knowledge graph, there exists a problem of missing or incorrect relationship connection with the specific entities. Numerous studies have proposed learning methods using artificial neural networks based on natural language embedding to solve the problems of the incomplete knowledge graph. Various knowledge graph completion systems are being studied using these methods. In this paper, a system that infers missing knowledge using specific queries and knowledge graphs is proposed. First, a topic is automatically extracted from a query, and topic embedding is obtained from the knowledge graph embedding module. Next, a new triple is inferred by learning the relationship between the topic from the knowledge graph and the query by using Query embedding and knowledge graph embedding. Through this method, the missing knowledge was inferred and the predicate embedding of the knowledge graph related to a specific query was used for good performance. Also, an experiment was conducted using the MetaQA dataset to prove the better performance of the proposed method compared with the existing methods. For the experiment, we used a knowledge graph having movies as a domain. Based on the assumption of the entire knowledge graph and the missing knowledge graph, we experimented on the knowledge graph in which 50% of the triples were randomly omitted. Apparently, better performance than the existing method was obtained.
Accuracy Evaluation Method for Image Classification Deep Learning Model considering Potentially Misclassified Data
Young-Woo Lee, Min-Ju Song, Heung-Seok Chae
http://doi.org/10.5626/JOK.2021.48.6.657
Image classification deep learning models have a problem of misclassifying the type of image when the image is modified. As the existing accuracy evaluation methods do not take into account images that can potentially be misclassified due to modification of the image, it becomes to trust the evaluation result even if the type of image is accurately classified. In this study, we have proposed a method for evaluating the accuracy of image classification deep learning models considering the potentially misclassified data. We have measured the boundary cost to identify potentially misclassified data for each model and data set and identified potentially misclassified data based on the boundary cost. Also, we have measured the accuracy considering the potentially misclassified data. As a result of identifying potentially misclassified data, 0.1～4.2% of the data were identified as potentially misclassified data, of which approximately 18～60% were actually misclassified. As a result of the accuracy evaluation, it was estimated that the higher-accuracy model was more robust to image modification, and the lower-accuracy model was less robust to image modification.
Smart Contract Weakness Analyzer Based on Concolic Testing
http://doi.org/10.5626/JOK.2021.48.6.668
Ethereum is a blockchain-based cryptocurrency platform that provides a Turing complete language, Solidity, which can be used to develop smart contracts for various applications. This paper present an analyzer that finds security weaknesses in smart contracts using the concolic testing framework. Concolic testing, which combines symbolic execution and testing, is more efficient than symbolic execution while retaining no false positiveness which is absent in static analysis. Also, the analyzer reflects actual execution context to the maximum extent possible using the Ethereum execution environment, the Geth testnet. The analyzer detects integer overflow and unhandled exception weakness. Also, this paper presents performance test results in comparison with a well known smart contract symbolic execution framework, Manticore.
An Explainable Knowledge Completion Model Using Explanation Segments
Min-Ho Lee, Wan-Gon Lee, Batselem Jagvaral, Young-Tack Park
http://doi.org/10.5626/JOK.2021.48.6.680
Recently, a large number of studies that used deep learning have been conducted to predict new links in incomplete knowledge graphs. However, link prediction using deep learning has a major limitation as the inferred results cannot be explained. We propose a high-utility knowledge graph prediction model that yields explainable inference paths supporting the inference results. We define paths to the object from the knowledge graph using a path ranking algorithm and define them as the explanation segments. Then, the generated explanation segments are embedded using a Convolutional neural network (CNN) and a Bidirectional Long short-term memory (BiLSTM). The link prediction model is then trained by applying an attention mechanism, based on the calculation of the semantic similarity between the embedded explanation segments and inferred candidate predicates to be inferred. The explanation segment suitable for link prediction explanation is selected based on the measured attention scores. To evaluate the performance of the proposed method, a link prediction comparison experiment and an accuracy verification experiment are performed to measure the proportion of the explanation segments suitable to explain the link prediction results. We used the benchmark datasets NELL-995, FB15K-237, and countries for the experiment, and accuracy verification experiments showed the accuracies of 89%, 44%, and 97%, respectively. Compared with the existing method, the NELL-995, FB15K-237 data exhibited 35%p and 21%p higher performance on average.
Topic Centric Korean Text Summarization using Attribute Model
Su-Hwan Yoon, A-Yeong Kim, Seong-Bae Park
http://doi.org/10.5626/JOK.2021.48.6.688
Abstractive summarization takes original text as an input and generates a summary containing the core-information about the original text. The abstractive summarization model is mainly designed by the Sequence-to-Sequence model. To improve quality as well as coherence of summary, the topic-centric methods which contain the core information of the original text are recently proposed. However, the previous methods perform additional training steps which make it difficult to take advantage of the pre-trained language model. This paper proposes a topic-centric summarizer that can reflect topic words to a summary as well as retain the characteristics of language model by using PPLM. The proposed method does not require any additional training. To prove the effectiveness of the proposed summarizer, this paper performed summarization experiments with Korean newspaper data.
A Comparative Study on the Performance of Named Entity Recognition in Materials and Chemistry Fields through Multiple Embedding Combination Based on a Pre-trained Neural Network Language Model
Myunghoon Lee, Hyeonho Shin, Hong-Woo Chun, Jae-Min Lee, Taehyun Ha, Sung-Pil Choi
http://doi.org/10.5626/JOK.2021.48.6.696
Recently, with the rapid development of materials and chemistry fields, the academic literature has increased exponentially. Accordingly, studies are being conducted to extract meaningful information from the existing accumulated data, and Named Entity Recognition (NER) is being utilized as one of the methodologies. NER in materials and chemistry fields is a task of extracting standardized entities such as materials, material property information, and experimental conditions from academic literature and classifying types of the entities. In this paper, we studied the NER in materials and chemistry fields using a combination of embedding and a Bi-direction LSTM-CRF model with an existing published language model without pre-training a neural network language model. As a result, we found the best performing embedding combinations and analyzed their performance. Additionally, the pre-trained language model was used as a NER model to compare performance through fine-tuning. The process showed that the use of a public pre-trained language model for embedding combinations could derive meaningful results in NER in the materials and chemistry fields.
Aircraft Reinforcement Learning using Curriculum Learning
Jung Ho Bae, Yun-Seong Kang, Sukmin Yoon, Yong-Duk Kim, Sungho Kim
http://doi.org/10.5626/JOK.2021.48.6.707
Diverse algorithms have been proposed to control unmanned aircrafts. However, they have limitations such as long exploration time and/or unclear behavior rules. To overcome the drawbacks and for efficient training, we propose an aircraft control technique using deep reinforcement learning applying antenna train angle (ATA) based curriculum learning. To validate the effectiveness of the proposed technique, we constructed a 3D simulation environment adapting a 6-DOF aircraft point model and conducted training with an initial setting of two fighters in the neutral position situation where they are looking back. The results showed that the proposed technique can achieve the goal of ATA 180° when the fighters are looking back without adding supplemental reward functions, while the deep reinforcement learning (DRL) without ATA curriculum could not succeed the learning over ATA 60° in a limited training time.
AttDRP: Attention Mechanism-based Model for Anti-Cancer Drug Response Prediction
Jonghwan Choi, Sangmin Seo, Sanghyun Park
http://doi.org/10.5626/JOK.2021.48.6.713
Resistance to anti-cancer drugs makes chemotherapy ineffective for cancer patients. Drug resistance is caused by genetic alterations in cancer cells. Many studies have investigated drug responses of diverse cancer cell lines to various anti-cancer drugs to understand drug response mechanisms. Existing studies have proposed machine learning models for drug response prediction to find effective anti-cancer drugs. However, currently there are no models to learn the relationship between anticancer drugs and genes to improve the prediction accuracy. In this paper, we proposed a predictive model AttDRP that could identify important genes associated with anti-cancer drugs and predict drug responses based on identified genes. AttDRP exhibited better predictive accuracy than existing models and we found that the attention scores of AttDRP could be effective tools to analyze molecular structures of anticancer drugs. We hope that our proposed method would contribute to the development of precision medicine for effective chemotherapy. Resistance to anti-cancer drugs makes chemotherapy ineffective for cancer patients. Drug resistance is caused by genetic alterations in cancer cells. Many studies have investigated drug responses of diverse cancer cell lines to various anti-cancer drugs to understand drug response mechanisms. Existing studies have proposed machine learning models for drug response prediction to find effective anti-cancer drugs. However, currently there are no models to learn the relationship between anticancer drugs and genes to improve the prediction accuracy. In this paper, we proposed a predictive model AttDRP that could identify important genes associated with anti-cancer drugs and predict drug responses based on identified genes. AttDRP exhibited better predictive accuracy than existing models and we found that the attention scores of AttDRP could be effective tools to analyze molecular structures of anticancer drugs.
Analysis of Limits in Applying AP-QoS-based Wi-Fi Slicing for Real-Time Systems
Jin Hyun Kim, Hyonyoung Choi, Gangjin Kim, Yundo Choi, Tae-Won Ban, Se-Hoon Kim
http://doi.org/10.5626/JOK.2021.48.6.723
Network slicing is a new network technology that guarantees the quality of network services according to application services or user’s types. Wi-Fi, IEEE 802.11-based LAN, is the mostly popularly used short-range wireless network and has been continually attracting more and more from users. Recently, the use of Wi-Fi by safety critical IoT devices, such as medical devices, has been drastically increasing. Moreover, enterprises require network slicing of Wi-Fi to introduce the provision of prioritized QoS of Wi-Fi depending on the service type of customer. This paper presents the analysis of the limits and difficulties in applying AP-QoS-based network slicing for hard real-time systems that demand temporal deterministic streaming services. In this paper, we have defined a formal framework to analyze QoS-providing IEEE 802.11e Enhanced Distributed Coordination Access and provide the worst-case streaming scenarios and thereby demonstrated why the temporal determinism of network streaming is broken. In addition, simulation results of AP-QoS-based network slicing using NS-3 are presented to show the limits and difficulties of the network slicing. Moreover, we present Wi-Fi network slicing techniques based on EDCA of AP-QoS for real-time systems through our technical report referenced in this paper.
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