发布求助
文献互助 智能选刊 最新文献

2021 13th International Conference on Knowledge and Systems Engineering (KSE)最新文献

筛选
英文 中文
Improving the Readability of Unformatted Text using Multitask Attention Networks 使用多任务注意网络提高未格式化文本的可读性
2021 13th International Conference on Knowledge and Systems Engineering (KSE) Pub Date : 2021-11-10 DOI: 10.1109/KSE53942.2021.9648633
V. Phan, Minh-Tien Nguyen, L. Bui, Phong Dao Ngoc
{"title":"Improving the Readability of Unformatted Text using Multitask Attention Networks","authors":"V. Phan, Minh-Tien Nguyen, L. Bui, Phong Dao Ngoc","doi":"10.1109/KSE53942.2021.9648633","DOIUrl":"https://doi.org/10.1109/KSE53942.2021.9648633","url":null,"abstract":"Unformatted text is a big obstacle to human reading and degrades the performance of many downstream language understanding tasks. To improve the readability, this paper proposes a multitask deep neural model to restore format standards including punctuation and capitalization. Unlike prior research which usually solved a single task or many tasks separately, our model employs multitask learning to simultaneously perform the restoration tasks. The model consists of a backbone network to learn language features, and attention-based predictors for the two tasks. To find the efficient encoding method for unformatted text, we analyze the model behaviour with different backbone architectures such as convolutional neural networks (CNN), unidirectional and bidirectional recurrent-based networks. The model is validated on two Vietnamese datasets and integrated into an automatic speech recognition (ASR) system. The experiments show the promising results for both restoration tasks and the applicability of our model.","PeriodicalId":130986,"journal":{"name":"2021 13th International Conference on Knowledge and Systems Engineering (KSE)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127314984","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Investigating Protection of Deep Learning Visual Features on ECB Encrypted Images ECB加密图像的深度学习视觉特征保护研究
2021 13th International Conference on Knowledge and Systems Engineering (KSE) Pub Date : 2021-11-10 DOI: 10.1109/KSE53942.2021.9648821
Kasidit Chunhachatchawhankhun, Prarinya Siritanawan, Karin Sumongkayothin, K. Kotani
{"title":"Investigating Protection of Deep Learning Visual Features on ECB Encrypted Images","authors":"Kasidit Chunhachatchawhankhun, Prarinya Siritanawan, Karin Sumongkayothin, K. Kotani","doi":"10.1109/KSE53942.2021.9648821","DOIUrl":"https://doi.org/10.1109/KSE53942.2021.9648821","url":null,"abstract":"In this paper, we demonstrate that images encrypted with Advanced Encryption Standard (AES) in Electronic Code Book (ECB) mode retain some local properties of the original images that Deep Neural Networks (DNNs) can detect these properties and perform classification tasks on this encrypted data. The experiment with the ECB encrypted MNIST handwritten digit dataset revealed that models trained on this dataset have an accuracy of around 80%. It also demonstrated that the model trained using one secret key does not work with other secret keys or the original dataset; the prediction accuracy plummeted to less than 10%. As a result, malicious users who do not know the secret keys will find the model inefficient, and it may be difficult to manipulate or change the prediction results.","PeriodicalId":130986,"journal":{"name":"2021 13th International Conference on Knowledge and Systems Engineering (KSE)","volume":"68 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129812472","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
Attention-Based Deep Learning Model for Aspect Classification on Vietnamese E-commerce Data 基于注意力的越南电子商务数据方面分类深度学习模型
2021 13th International Conference on Knowledge and Systems Engineering (KSE) Pub Date : 2021-11-10 DOI: 10.1109/KSE53942.2021.9648690
Ngoc-Tu Nguyen, Trong-Dat Nguyen, Duy-Cat Can, Mai-Vu Tran, Ha Luu Manh, Hoang-Quynh Le
{"title":"Attention-Based Deep Learning Model for Aspect Classification on Vietnamese E-commerce Data","authors":"Ngoc-Tu Nguyen, Trong-Dat Nguyen, Duy-Cat Can, Mai-Vu Tran, Ha Luu Manh, Hoang-Quynh Le","doi":"10.1109/KSE53942.2021.9648690","DOIUrl":"https://doi.org/10.1109/KSE53942.2021.9648690","url":null,"abstract":"This article introduces methods for applying Deep Learning in identifying aspects from written commentaries on Shopee e-commerce sites. The used datasets are two sets of Vietnamese consumers' comments about purchased products in two domains. Words and sentences will be performed as vectors, or characteristic matrices through language models such as one-hot, fastText, PhoBERT. We then used Convolutional Neural Network (CNN) and the Fully Connected Neural Network (Multilayer perceptron - MLP) to learn the aspects which are mentioned in the comments. Experimental results showed that our team's methods achieved much better results than traditional learning algorithm using other word-level vectors such as SVM, Naïve Bayes, etc.","PeriodicalId":130986,"journal":{"name":"2021 13th International Conference on Knowledge and Systems Engineering (KSE)","volume":"28 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121999910","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 2
Detection of Wrong Direction Vehicles on Two-Way Traffic 双向交通错误方向车辆的检测
2021 13th International Conference on Knowledge and Systems Engineering (KSE) Pub Date : 2021-11-10 DOI: 10.1109/KSE53942.2021.9648579
Pintusorn Suttiponpisarn, C. Charnsripinyo, Sasiporn Usanavasin, Hiro Nakahara
{"title":"Detection of Wrong Direction Vehicles on Two-Way Traffic","authors":"Pintusorn Suttiponpisarn, C. Charnsripinyo, Sasiporn Usanavasin, Hiro Nakahara","doi":"10.1109/KSE53942.2021.9648579","DOIUrl":"https://doi.org/10.1109/KSE53942.2021.9648579","url":null,"abstract":"In 2018, around 22,000 people were killed in road accidents in Thailand and more than 80% of road accidents involved motorcycles. There are several reasons and one of them is motorcycles driving in the wrong direction. Our paper proposes a system for road surveillance to detect and track moving vehicle's direction on two-way traffic roads from CCTV viewpoint using deep learning where the roadside noise is filtered out. The system can determine whether the vehicles drive in the correct direction in their lane. In this case, we will focus on motorcycles. To set up the system, there are two main steps: draw the area of interest and validate the correct direction. In the direction validation process, we detect and track vehicles by using YOLOv4-tiny and Deep SORT tracking algorithm. The system returns a satisfactory and accurate result for detecting direction without noise, such as moving objects on the sidewalk or moving vehicles in an unwanted area.","PeriodicalId":130986,"journal":{"name":"2021 13th International Conference on Knowledge and Systems Engineering (KSE)","volume":"120 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123440940","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 2
A Combination of DWT and QR Decomposition for Color Image Watermarking 基于DWT和QR分解的彩色图像水印
2021 13th International Conference on Knowledge and Systems Engineering (KSE) Pub Date : 2021-11-10 DOI: 10.1109/KSE53942.2021.9648714
Phuong Thi Nha, Ta Minh Thanh
{"title":"A Combination of DWT and QR Decomposition for Color Image Watermarking","authors":"Phuong Thi Nha, Ta Minh Thanh","doi":"10.1109/KSE53942.2021.9648714","DOIUrl":"https://doi.org/10.1109/KSE53942.2021.9648714","url":null,"abstract":"For the purpose of protecting copyright of digital images, many published articles indicate that combining DWT with matrix decomposition can improve robustness of watermark. In this paper, a hybrid image watermarking scheme is represented by using a combination of 1-level $DWT$ transform and $QR$ decomposition. The first, Haar transform is applied on the channel B of the host color image and $LL$ sub-band is partitioned in $4 times 4$ non-overlapping blocks. After that, $QR$ factorization is executed on each block where $Q$ matrix and $R$ matrix are calculated separately by our proposal. Watermark bits are embedded into the first element of $R$ matrix instead of the whole row as previous published articles to improve the quality of watermarked image. In addition, the original watermark image is permuted by Arnold transform and it helps the proposed method is more secure. In this study, we also perform experiments to select the most suitable sub-band for embedding information. The experimental results show that this combination gives a better invisibility of watermarked image which is compared to the $QR$ decomposition based approach. Furthermore, the proposed algorithm is more effective to against many common attacks.","PeriodicalId":130986,"journal":{"name":"2021 13th International Conference on Knowledge and Systems Engineering (KSE)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126850461","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 1
Kodiak@ALQAC2021: Deep Learning for Vietnamese Legal Information Processing Kodiak@ALQAC2021:越南法律信息处理的深度学习
2021 13th International Conference on Knowledge and Systems Engineering (KSE) Pub Date : 2021-11-10 DOI: 10.1109/KSE53942.2021.9648744
Dinh-Truong Do
{"title":"Kodiak@ALQAC2021: Deep Learning for Vietnamese Legal Information Processing","authors":"Dinh-Truong Do","doi":"10.1109/KSE53942.2021.9648744","DOIUrl":"https://doi.org/10.1109/KSE53942.2021.9648744","url":null,"abstract":"In this paper, we propose deep learning based methods for addressing the problems of legal text processing in the Automated Legal Question Answering Competition (ALQAC 2021). The competition consists of three challenging tasks based on well-known statute laws in Vietnamese and Thai Language. We participated in three tasks related to the Vietnamese statute law, including the legal document retrieval (Task 1), the legal textual entailment (Task 2), and the legal question answering (Task 3). In Task 1, we combine semantic and lexical scores to identify relevant articles. In Tasks 2&3, we fine-tune pretrained models for the Vietnamese language in order to classify proper labels. The experimental results demonstrate both the difficulties and potentials associated with these approaches.","PeriodicalId":130986,"journal":{"name":"2021 13th International Conference on Knowledge and Systems Engineering (KSE)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131627968","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 2
An Embedding Method for Sentiment Classification across Multiple Languages 一种跨语言情感分类的嵌入方法
2021 13th International Conference on Knowledge and Systems Engineering (KSE) Pub Date : 2021-11-10 DOI: 10.1109/KSE53942.2021.9648795
Le-Tuan Duy Nguyen, Ngoc Dung Nguyen, Khac-Hoai Nam Bui
{"title":"An Embedding Method for Sentiment Classification across Multiple Languages","authors":"Le-Tuan Duy Nguyen, Ngoc Dung Nguyen, Khac-Hoai Nam Bui","doi":"10.1109/KSE53942.2021.9648795","DOIUrl":"https://doi.org/10.1109/KSE53942.2021.9648795","url":null,"abstract":"Embedding methods are feature representations of words, which are able to capture both semantic and syntactic information from contexts. However, existing embedding methods for learning context are typically unable to capturing sufficient sentiment information. In this study, we conduct an investigation on how to improve the performance of sentiment classification using sentiment embedding approach. Particularly, we first present a new word embedding method based on a supervised method to capture the semantic sentiment information. Then, Deep Learning models, by combining different architecture such as Recurrent Neural Network (RNN), Convolutional Neural Network (CNN), and attention mechanisms are developed for improving the performance of the sentiment classification. The evaluation on well-known bench-mark datasets with different languages (i.e., English and Vietnamese) indicates the promising results of our method.","PeriodicalId":130986,"journal":{"name":"2021 13th International Conference on Knowledge and Systems Engineering (KSE)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132009569","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 1
Simulation of precision feeding systems for swine 猪精确饲养系统的仿真
2021 13th International Conference on Knowledge and Systems Engineering (KSE) Pub Date : 2021-11-10 DOI: 10.1109/KSE53942.2021.9648760
L. Pham, H. Nguyen-Ba, Hoai Son Nguyen, Huy-Ham Le
{"title":"Simulation of precision feeding systems for swine","authors":"L. Pham, H. Nguyen-Ba, Hoai Son Nguyen, Huy-Ham Le","doi":"10.1109/KSE53942.2021.9648760","DOIUrl":"https://doi.org/10.1109/KSE53942.2021.9648760","url":null,"abstract":"Precision livestock farming has become an inevitable trend for agricultural industry in the world. In that field, precision feeding is widely acknowledged because of its potential to reduce feed costs, environmental footprint and to improve animal health and welfare. Precision feeding involves modern multidisciplinary technologies such as information technology, mechanics, electronics, automation, etc. Such a system consists of automatic troughs linked to a computer system to exploit data collected from the individual animals (e.g. body weight, feed intake and feeding behaviour), and/or from ambient sensors. Data is processed and analysed based on mathematical models to make predictions, warnings for farmers or to formulate diets that fit requirements of each individual animal at each production period. However, implementing such a system often requires high investment, which may go beyond the capabilities of smallholders and small/medium laboratories. Furthermore, the risk of implementing by design but not conforming to reality is very high. To avoid this problem, we introduce an agent-based modelling approach to simulate precision feeding systems for swine. Two simulation experiments were conducted to provide predictions about the growth of individual pigs and the usefulness of precision feeding systems over classic feeding models.","PeriodicalId":130986,"journal":{"name":"2021 13th International Conference on Knowledge and Systems Engineering (KSE)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132145650","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 1
An incremental ensemble learning system for Vietnamese e-commerce product classification 越南电子商务产品分类的增量集成学习系统
2021 13th International Conference on Knowledge and Systems Engineering (KSE) Pub Date : 2021-11-10 DOI: 10.1109/KSE53942.2021.9648642
Linh Nguyen Tran Ngoc, Vu Hong Quan, Le Hoang Ngan, Tran Duy Phu, Hoang-Quynh Le
{"title":"An incremental ensemble learning system for Vietnamese e-commerce product classification","authors":"Linh Nguyen Tran Ngoc, Vu Hong Quan, Le Hoang Ngan, Tran Duy Phu, Hoang-Quynh Le","doi":"10.1109/KSE53942.2021.9648642","DOIUrl":"https://doi.org/10.1109/KSE53942.2021.9648642","url":null,"abstract":"With the booming of e-commerce platforms, text classification models play an increasingly important role in businesses. Major challenges that businesses would face include dataset imbalance, continuously added data, language specificity and product specificity. In this paper, we propose a scalable incremental machine learning system for industrial-scale deployment in real-world business. The system also includes steps to optimize e-commerce product specifics. The proposal tactics including keyword dictionary mapping, sampling technique and ensemble learning delivered better performance when compared to models without them. Our experiments also showed that minibatch SVM produced good results and might be considerable in a lighter system.","PeriodicalId":130986,"journal":{"name":"2021 13th International Conference on Knowledge and Systems Engineering (KSE)","volume":"63 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122983006","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 1
A Hierarchical Long Short-Term Memory Encoder-Decoder Model for Abstractive Summarization 面向抽象摘要的长短期记忆分层编码器模型
2021 13th International Conference on Knowledge and Systems Engineering (KSE) Pub Date : 2021-11-10 DOI: 10.1109/KSE53942.2021.9648836
Khuong Nguyen-Ngoc, Anh-Cuong Le, Viet-Ha Nguyen
{"title":"A Hierarchical Long Short-Term Memory Encoder-Decoder Model for Abstractive Summarization","authors":"Khuong Nguyen-Ngoc, Anh-Cuong Le, Viet-Ha Nguyen","doi":"10.1109/KSE53942.2021.9648836","DOIUrl":"https://doi.org/10.1109/KSE53942.2021.9648836","url":null,"abstract":"Abstractive summarization is the task of generating concise summary of a source text, which is a challenging problem in Natural Language Processing (NLP). Many recent studies have relied on encoder-decoder sequence-to-sequence deep neural networks to solve this problem. However, most of these models treat the input as a sequence of words at the same level during the encoding process. This will make the encoding inefficient, especially for long input texts. Addressing this issue, in this paper we propose a model to encode text in a hierarchical manner, which helps to encode information in a way that is consistent with the nature of the text: the text is synthesized from sentences, and each sentence is synthesized from words. Our proposed model is based on Long Short Term Memory model that we called HLSTM (Hierarchical Long Short Term Memory) and applied to the problem of abstractive summarization. We conducted extensive experiments on the two most popular corpora (Gigaword and Amazon Review) and obtain significant improvements in comparison with the baseline models.","PeriodicalId":130986,"journal":{"name":"2021 13th International Conference on Knowledge and Systems Engineering (KSE)","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131563202","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
引用次数: 0
首页 上一页 下一页 尾页
0
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
提示
您的信息不完整，为了账户安全，请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
请完成安全验证×
相关产品
×
本文献相关产品
联系我们：info@booksci.cn Book学术提供免费学术资源搜索服务，方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。 Copyright © 2023 布克学术 All rights reserved.
京ICP备2023020795号-1
ghs 京公网安备 11010802042870号
Book学术文献互助
Book学术文献互助群
群 号：604180095
Book学术官方微信