Question-answer platforms are trending over the internet nowadays. Clustering questions which ask the same question is a challenging problem faced by such platforms. The authors will discuss and implement techniques that will tell them whether a given pair of questions are asking the same question. They will do this by finding the semantic relationship between the questions. Manhattan Long short-term memory model with Area Under ROC Curve 0.81 is the model which performs the best among all the trained and pre-trained models which they will be discussing in this paper.
{"title":"Learning Question Similarity","authors":"Pooja Bihani, Ashay Walke","doi":"10.1145/3380688.3380713","DOIUrl":"https://doi.org/10.1145/3380688.3380713","url":null,"abstract":"Question-answer platforms are trending over the internet nowadays. Clustering questions which ask the same question is a challenging problem faced by such platforms. The authors will discuss and implement techniques that will tell them whether a given pair of questions are asking the same question. They will do this by finding the semantic relationship between the questions. Manhattan Long short-term memory model with Area Under ROC Curve 0.81 is the model which performs the best among all the trained and pre-trained models which they will be discussing in this paper.","PeriodicalId":414793,"journal":{"name":"Proceedings of the 4th International Conference on Machine Learning and Soft Computing","volume":"102 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115327970","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}
T. Ngo, P. V. Dang, Thanh Vo Nhu, Le Hoai Nam, Le Hung Toan Do, L. A. Doan
Medical sector is one of the important domains in current society. This paper focuses on the patient treatment processes, in case of requiring prosthesis implantation. The specificity of such a process is that it makes connections between two lifecycles belonging to medical and engineering domains respectively. This implies several collaborative actions between stakeholders from heterogeneous disciplines. However, several problems of communication and knowledge sharing may occur because of the variety of semantic used and the specific business practices in each domain. In this context, this paper is interested in the potential of knowledge engineering and product lifecycle management approaches to cope with the above problems. To do so, a conceptual framework is proposed for the analysis of links between the disease (medical domain) and the prosthesis (engineering domain) lifecycles. Based on this analysis, a combined KM-PLM based approach is proposed. The application of the proposition is demonstrated through an implementation of useful function in the AUDROS PLM software.
{"title":"Towards an Effective Solution for Medical Treatment Process based on Product Lifecycle Management","authors":"T. Ngo, P. V. Dang, Thanh Vo Nhu, Le Hoai Nam, Le Hung Toan Do, L. A. Doan","doi":"10.1145/3380688.3380700","DOIUrl":"https://doi.org/10.1145/3380688.3380700","url":null,"abstract":"Medical sector is one of the important domains in current society. This paper focuses on the patient treatment processes, in case of requiring prosthesis implantation. The specificity of such a process is that it makes connections between two lifecycles belonging to medical and engineering domains respectively. This implies several collaborative actions between stakeholders from heterogeneous disciplines. However, several problems of communication and knowledge sharing may occur because of the variety of semantic used and the specific business practices in each domain. In this context, this paper is interested in the potential of knowledge engineering and product lifecycle management approaches to cope with the above problems. To do so, a conceptual framework is proposed for the analysis of links between the disease (medical domain) and the prosthesis (engineering domain) lifecycles. Based on this analysis, a combined KM-PLM based approach is proposed. The application of the proposition is demonstrated through an implementation of useful function in the AUDROS PLM software.","PeriodicalId":414793,"journal":{"name":"Proceedings of the 4th International Conference on Machine Learning and Soft Computing","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128098803","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}
Sentimental data is now a trend can be generally considered into two key types mainly facts and feelings. Facts are unbiased expressions around entities, actions, and their belongings. The thoughts of estimation in terms of sentiments are very extensive. In this paper, the main focus is given on the opinion terminologies that carry positive or negative thoughts. These thoughts are considered as sentiments. Plentiful work is done already using text processing in terms of mining of the information and recovery of the data. It is done using clustering approaches, mining of the text and other various text mining tasks but very less work is in handling of opinions in the medical field. Yet, sentiments are so imperative in the medical field to make decisions. The dataset on which the processing is done is the digital retinal DRIVE dataset was taken with 8-BPC (bits per color level) at 768 × 584 pixels. So this paper put light on the efficient approach for sentiment analysis using normalization and feature extraction for high classification rates and the simulation environment is used as MATLAB for development purpose.
{"title":"An Effectual Sentiment Analysis for High Classification Rates Using Medical Image Processing","authors":"G. Jaitly, Manoj Kapil","doi":"10.1145/3380688.3380714","DOIUrl":"https://doi.org/10.1145/3380688.3380714","url":null,"abstract":"Sentimental data is now a trend can be generally considered into two key types mainly facts and feelings. Facts are unbiased expressions around entities, actions, and their belongings. The thoughts of estimation in terms of sentiments are very extensive. In this paper, the main focus is given on the opinion terminologies that carry positive or negative thoughts. These thoughts are considered as sentiments. Plentiful work is done already using text processing in terms of mining of the information and recovery of the data. It is done using clustering approaches, mining of the text and other various text mining tasks but very less work is in handling of opinions in the medical field. Yet, sentiments are so imperative in the medical field to make decisions. The dataset on which the processing is done is the digital retinal DRIVE dataset was taken with 8-BPC (bits per color level) at 768 × 584 pixels. So this paper put light on the efficient approach for sentiment analysis using normalization and feature extraction for high classification rates and the simulation environment is used as MATLAB for development purpose.","PeriodicalId":414793,"journal":{"name":"Proceedings of the 4th International Conference on Machine Learning and Soft Computing","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114983788","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}
One of the keys to enable chatbots to communicate with human in a more natural way is the ability to handle long and complex user's utterances. In order to achieve this goal, we propose to integrate the Sentence Boundary Detection (SBD) module into the chatbot architecture, whose role is to take as input a user's utterance from an automatic speech recognition device, in which sentence boundaries are not available, and output the corresponding list of punctuated sentences for downstream modules such as Intent Detection, Topic Classification, Sentiment Analysis, Named Entity Recognition, as well as Coreference Recognition. To address the SBD task, we reformulate it as a sequence labeling task. In this way, both deep neural network models (e.g., Bi-directional Long Short-Term Memory, Convolutional Neural Network) and structured prediction models (e.g., Hidden Markov Model, Maximum Entropy Model, Conditional Random Field) can be leveraged. After reformulating the SBD task, we built a hybrid deep neural network model and achieved good performance on both CornellMovie-Dialog and DailyDialog datasets.
{"title":"Sequence Labeling Approach to the Task of Sentence Boundary Detection","authors":"T. A. Le","doi":"10.1145/3380688.3380703","DOIUrl":"https://doi.org/10.1145/3380688.3380703","url":null,"abstract":"One of the keys to enable chatbots to communicate with human in a more natural way is the ability to handle long and complex user's utterances. In order to achieve this goal, we propose to integrate the Sentence Boundary Detection (SBD) module into the chatbot architecture, whose role is to take as input a user's utterance from an automatic speech recognition device, in which sentence boundaries are not available, and output the corresponding list of punctuated sentences for downstream modules such as Intent Detection, Topic Classification, Sentiment Analysis, Named Entity Recognition, as well as Coreference Recognition. To address the SBD task, we reformulate it as a sequence labeling task. In this way, both deep neural network models (e.g., Bi-directional Long Short-Term Memory, Convolutional Neural Network) and structured prediction models (e.g., Hidden Markov Model, Maximum Entropy Model, Conditional Random Field) can be leveraged. After reformulating the SBD task, we built a hybrid deep neural network model and achieved good performance on both CornellMovie-Dialog and DailyDialog datasets.","PeriodicalId":414793,"journal":{"name":"Proceedings of the 4th International Conference on Machine Learning and Soft Computing","volume":"212 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123735886","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}
Thanh Vo Nhu, D. Vinh, Le Hoai Nam, Ngo Thanh Nghi, Doan Le Anh
The wheeled mobile robots have been widely applied in daily applications for its simplicity, robustness, stability and low-cost of manufacturing. Typically, PID controller is implemented to drive a mobile robot. However, the performance of the robot is greatly depended on the PID control parameters are tuned and the robustness of the mechanical system. Since the analog controllers were replaced by digital controller, PID control parameters could be automatically tuned to adapt the variation of the system parameters and operating condition. In this manuscript, a mathematical model is developed for a line following mobile robotic system which is used as a restaurant serving robot. The serving robot is controlled by an adaptive PID controller using Q-learning algorithm. The simulation and experimental results are compared to verify the advantage of the adaptive PID controller over classical PID controller.
{"title":"Reinforcement Q-learning PID Controller for a Restaurant Mobile Robot with Double Line-Sensors","authors":"Thanh Vo Nhu, D. Vinh, Le Hoai Nam, Ngo Thanh Nghi, Doan Le Anh","doi":"10.1145/3380688.3380718","DOIUrl":"https://doi.org/10.1145/3380688.3380718","url":null,"abstract":"The wheeled mobile robots have been widely applied in daily applications for its simplicity, robustness, stability and low-cost of manufacturing. Typically, PID controller is implemented to drive a mobile robot. However, the performance of the robot is greatly depended on the PID control parameters are tuned and the robustness of the mechanical system. Since the analog controllers were replaced by digital controller, PID control parameters could be automatically tuned to adapt the variation of the system parameters and operating condition. In this manuscript, a mathematical model is developed for a line following mobile robotic system which is used as a restaurant serving robot. The serving robot is controlled by an adaptive PID controller using Q-learning algorithm. The simulation and experimental results are compared to verify the advantage of the adaptive PID controller over classical PID controller.","PeriodicalId":414793,"journal":{"name":"Proceedings of the 4th International Conference on Machine Learning and Soft Computing","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126975905","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}
Sitting improperly can impact on the amount of pressure put on the back of the body, causing the spine to degrade prematurely. Therefore, this paper proposed a method of computing sitting angles and the creation of a mobile application to alert users when they are sitting improperly. First, a digital protractor and an accelerometer-equipped with a smartphone were employed to gather data. Next, regression equations derived from the collected data were applied to compute the angles between the upper and lower body as well as the angles on the left and right sides of the upper body while sitting. The computed angles were used to analyze the sitting posture. To expedite the testing process, an Android Sensor Fusion application was used to stream the accelerometer data over a local network through a router to a PC-based program. Finally, a mobile application was created. The results showed that the proposed method could be used to obtain sitting angles and alert users when they were sitting improperly.
{"title":"Developing a Mobile Application to Detect Improper Sitting Using Regression Analysis and an Accelerometer","authors":"R. Phoophuangpairoj, P. Charnkeitkong","doi":"10.1145/3380688.3380698","DOIUrl":"https://doi.org/10.1145/3380688.3380698","url":null,"abstract":"Sitting improperly can impact on the amount of pressure put on the back of the body, causing the spine to degrade prematurely. Therefore, this paper proposed a method of computing sitting angles and the creation of a mobile application to alert users when they are sitting improperly. First, a digital protractor and an accelerometer-equipped with a smartphone were employed to gather data. Next, regression equations derived from the collected data were applied to compute the angles between the upper and lower body as well as the angles on the left and right sides of the upper body while sitting. The computed angles were used to analyze the sitting posture. To expedite the testing process, an Android Sensor Fusion application was used to stream the accelerometer data over a local network through a router to a PC-based program. Finally, a mobile application was created. The results showed that the proposed method could be used to obtain sitting angles and alert users when they were sitting improperly.","PeriodicalId":414793,"journal":{"name":"Proceedings of the 4th International Conference on Machine Learning and Soft Computing","volume":"145 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124727075","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}
User identifications are in searching Online Social Networks (OSN) to find identical users among different social sites in many data sources (data integration, data enrichment, information retrieval,...). However, these user-unique attributes are difficult to obtain due to privacy issues. It is hard to identify users across multiple OSNs online. This paper has presented user's identification across multiple OSNs in order to develop searching engine for user identification. The proposed approach is designed to find by searching engine while accommodating User identifications in searching Online Social Networks (OSN). Experimental results demonstrate that our proposed approach achieves a significant improvement in term of performance accuracy.
{"title":"A Novel Approach using Context Matching Algorithm and Knowledge Inference for User Identification in Social Networks","authors":"H. Pham, Van Thai Nguyen","doi":"10.1145/3380688.3380708","DOIUrl":"https://doi.org/10.1145/3380688.3380708","url":null,"abstract":"User identifications are in searching Online Social Networks (OSN) to find identical users among different social sites in many data sources (data integration, data enrichment, information retrieval,...). However, these user-unique attributes are difficult to obtain due to privacy issues. It is hard to identify users across multiple OSNs online. This paper has presented user's identification across multiple OSNs in order to develop searching engine for user identification. The proposed approach is designed to find by searching engine while accommodating User identifications in searching Online Social Networks (OSN). Experimental results demonstrate that our proposed approach achieves a significant improvement in term of performance accuracy.","PeriodicalId":414793,"journal":{"name":"Proceedings of the 4th International Conference on Machine Learning and Soft Computing","volume":"92 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133785365","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}
The objective of this study is to use government procurement open data to inform Civil Society Organizations (CSOs) of the best possible strategies to optimize, and maximize their advocacy work. In this paper, pre-processing of the Department of Health Region (DOH) 5 procurement datasets from 2016- 2018 are conducted, collected from the official Philippine Government E-Procurement System (PhilGEPS) repository. The visualizations and calculations showed average prices of HIV-related commodities, procurement categories, average procurement timelines and patterns, procurement allocation per province, which CSOs, represented by Gayon Bicol, may use to improve their crafting of proposal and monitoring of service delivery in their HIV-related grassroots advocacy work in the province of Albay. The data also showed posting of Notice to Proceed in PhilGEPS by approximately 16 days earlier than Publication of Award, for the entire Department of Health (DOH) Region 5, and 108 days early on HIV- related transactions, which is not within the compliance standards as stated in the Government Procurement Reform Act (GPRA) law, thus, a point for improvement for the agencies concerned.
{"title":"OCDex Portal: A Data-Driven Approach in Analyzing Procurement Process of Community-Based HIV/AIDS Advocacy Related Items","authors":"Maria Jihan G. Sangil, Lany L. Maceda","doi":"10.1145/3380688.3380719","DOIUrl":"https://doi.org/10.1145/3380688.3380719","url":null,"abstract":"The objective of this study is to use government procurement open data to inform Civil Society Organizations (CSOs) of the best possible strategies to optimize, and maximize their advocacy work. In this paper, pre-processing of the Department of Health Region (DOH) 5 procurement datasets from 2016- 2018 are conducted, collected from the official Philippine Government E-Procurement System (PhilGEPS) repository. The visualizations and calculations showed average prices of HIV-related commodities, procurement categories, average procurement timelines and patterns, procurement allocation per province, which CSOs, represented by Gayon Bicol, may use to improve their crafting of proposal and monitoring of service delivery in their HIV-related grassroots advocacy work in the province of Albay. The data also showed posting of Notice to Proceed in PhilGEPS by approximately 16 days earlier than Publication of Award, for the entire Department of Health (DOH) Region 5, and 108 days early on HIV- related transactions, which is not within the compliance standards as stated in the Government Procurement Reform Act (GPRA) law, thus, a point for improvement for the agencies concerned.","PeriodicalId":414793,"journal":{"name":"Proceedings of the 4th International Conference on Machine Learning and Soft Computing","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132669533","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}
Nutritional status plays an important role in not only pregnancy outcomes but also neonatal health. One of efficient techniques to control the nutritional status is to estimate the energy expenditure. There are some approaches for estimating energy expenditure. However, they have limitations including high cost, relative complexity, trained personnel requirements, or locality. This study investigates in a system for data collection and analysis (IoH-Internet of Health) developing based on microservice architecture, and its application for energy expenditure estimation. The proposed system has a good ability to scale and integrate with other systems; the energy expenditure estimation is performed by using artificial intelligence. The experimental results have shown the promising results of the proposed system.
{"title":"Energy expenditure estimation based on artificial intelligence and microservice architecture","authors":"H. T. Huynh, H. Quan","doi":"10.1145/3380688.3380715","DOIUrl":"https://doi.org/10.1145/3380688.3380715","url":null,"abstract":"Nutritional status plays an important role in not only pregnancy outcomes but also neonatal health. One of efficient techniques to control the nutritional status is to estimate the energy expenditure. There are some approaches for estimating energy expenditure. However, they have limitations including high cost, relative complexity, trained personnel requirements, or locality. This study investigates in a system for data collection and analysis (IoH-Internet of Health) developing based on microservice architecture, and its application for energy expenditure estimation. The proposed system has a good ability to scale and integrate with other systems; the energy expenditure estimation is performed by using artificial intelligence. The experimental results have shown the promising results of the proposed system.","PeriodicalId":414793,"journal":{"name":"Proceedings of the 4th International Conference on Machine Learning and Soft Computing","volume":"127 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116793895","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}
Zhicheng Chen, Xiaobao Liu, Yanchao Yin, Hongbiao Lu
Aiming at the problem that fault text is difficult to be directly parsed and utilized, a fault knowledge extraction method is proposed based on deep learning method. Firstly, the characteristics of fault knowledge are analyzed, and a fault knowledge extraction model is established based on multi-layer neural network. Finally, the presented model is discussed comprehensively from the extraction accuracy, recall rate and F1 value, which proves the feasibility of the method. The unstructured text data is used to provide reference for fault diagnosis and prediction.
{"title":"Named Entity Recognition Method for Fault Knowledge based on Deep Learning","authors":"Zhicheng Chen, Xiaobao Liu, Yanchao Yin, Hongbiao Lu","doi":"10.1145/3380688.3380690","DOIUrl":"https://doi.org/10.1145/3380688.3380690","url":null,"abstract":"Aiming at the problem that fault text is difficult to be directly parsed and utilized, a fault knowledge extraction method is proposed based on deep learning method. Firstly, the characteristics of fault knowledge are analyzed, and a fault knowledge extraction model is established based on multi-layer neural network. Finally, the presented model is discussed comprehensively from the extraction accuracy, recall rate and F1 value, which proves the feasibility of the method. The unstructured text data is used to provide reference for fault diagnosis and prediction.","PeriodicalId":414793,"journal":{"name":"Proceedings of the 4th International Conference on Machine Learning and Soft Computing","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-01-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122953030","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：