Pub Date : 2022-12-10DOI: 10.1109/ICSAI57119.2022.10005391
Baosheng Yin, Wei Zhao
In the field of biomedical text mining, biomedical entity relation extraction is the core task to assist researchers in completing the text mining. However, the special characteristics of medical literature, such as long and complex syntax of medical text and a large number of overlapping relations, pose a challenge for biomedical entity relation extraction. In this paper, we propose and apply a Directed Graph Convolutional Network (D-GCN) to encode syntactic information based on the existing neural model as a backbone, thus enhancing the representational ability of the input sequence. Experiments on the evaluation data set showed that the framework could effectively enhance the baseline models based on LSTM, Transformer and pre-trained model BERT at three different levels. Compared with the baseline model PRGC, our model improved F1-score by 1.4% on average.
{"title":"Chinese Biomedical Entity Relation Extraction Based On Directed Graph Convolutional Network","authors":"Baosheng Yin, Wei Zhao","doi":"10.1109/ICSAI57119.2022.10005391","DOIUrl":"https://doi.org/10.1109/ICSAI57119.2022.10005391","url":null,"abstract":"In the field of biomedical text mining, biomedical entity relation extraction is the core task to assist researchers in completing the text mining. However, the special characteristics of medical literature, such as long and complex syntax of medical text and a large number of overlapping relations, pose a challenge for biomedical entity relation extraction. In this paper, we propose and apply a Directed Graph Convolutional Network (D-GCN) to encode syntactic information based on the existing neural model as a backbone, thus enhancing the representational ability of the input sequence. Experiments on the evaluation data set showed that the framework could effectively enhance the baseline models based on LSTM, Transformer and pre-trained model BERT at three different levels. Compared with the baseline model PRGC, our model improved F1-score by 1.4% on average.","PeriodicalId":339547,"journal":{"name":"2022 8th International Conference on Systems and Informatics (ICSAI)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131385947","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}
Pub Date : 2022-12-10DOI: 10.1109/ICSAI57119.2022.10005470
Xiaojing Zhang, Dongshuo Zhao, Yaran Li, Yudong Liu, Gang Hu
With the increasing intensity of competition in the current metrology testing market, building customer portrait is an effective way for metrology institutions to improve service levels to customers. This paper is based on the basic business data of a certain metrology institution. First, recency, frequency, monetary value model (RFM model), which is widely applied in customer relationship management, is improved. Further, it is combined with the business features of the metrology institution and used to build data feature engineering, which is closely related to the business data of the metrology institution and can reflect the data situation. Then, the data are analyzed through correlation test, standardized by Z-score, and clustered with three clustering algorithms, namely K-Means, DBSCAN, and AGNES, which are in SKLEARN database based on Python. After that, the clustering results are compared. In the clustering process, the elbow method and method for traversing the silhouette coefficient are used to determine the optimal value of the clustering algorithm. Finally, with the analysis of clustering results, the customers’ features of the metrology institution are signed and the customer portrait is built, which provides data analysis methods, tools and decision basis for the metrology institution to offer better services.
{"title":"Customer Portrait for Metrology Institutions Based on the Machine Learning Clustering Algorithm and the RFM Model","authors":"Xiaojing Zhang, Dongshuo Zhao, Yaran Li, Yudong Liu, Gang Hu","doi":"10.1109/ICSAI57119.2022.10005470","DOIUrl":"https://doi.org/10.1109/ICSAI57119.2022.10005470","url":null,"abstract":"With the increasing intensity of competition in the current metrology testing market, building customer portrait is an effective way for metrology institutions to improve service levels to customers. This paper is based on the basic business data of a certain metrology institution. First, recency, frequency, monetary value model (RFM model), which is widely applied in customer relationship management, is improved. Further, it is combined with the business features of the metrology institution and used to build data feature engineering, which is closely related to the business data of the metrology institution and can reflect the data situation. Then, the data are analyzed through correlation test, standardized by Z-score, and clustered with three clustering algorithms, namely K-Means, DBSCAN, and AGNES, which are in SKLEARN database based on Python. After that, the clustering results are compared. In the clustering process, the elbow method and method for traversing the silhouette coefficient are used to determine the optimal value of the clustering algorithm. Finally, with the analysis of clustering results, the customers’ features of the metrology institution are signed and the customer portrait is built, which provides data analysis methods, tools and decision basis for the metrology institution to offer better services.","PeriodicalId":339547,"journal":{"name":"2022 8th International Conference on Systems and Informatics (ICSAI)","volume":"85 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114290321","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}
Pub Date : 2022-12-10DOI: 10.1109/ICSAI57119.2022.10005357
Peng Yang, Weishun Xue, Chenqi Fan, Xiaojian Xue
with the growing concerns of the current network security, development of hybrid cryptosystem based on the combination of optical symmetric cryptography and public key cryptography has aroused considerable research interest. In optical symmetry system, a new method for encrypting optical information using Fresnel lenses is proposed. The symmetry system’s keys include the focal length (f) and propagation wavelength ($lambda$). Then, in public key cryptography, the key ($lambda$,J) is encrypted by the RSA algorithm. Finally, we have implemented and result have been obtained using 12-level digital Fresnel lens by COMSOL Multiphysics to study of how the optical image deciphered the information using encryption keys. In addition, the main challenges and future prospects for the further development of hybrid cryptosystem system are discussed.
{"title":"Promoting a Hybrid Cryptosystem System’s Security based on Fresnel lens and RSA Algorithm","authors":"Peng Yang, Weishun Xue, Chenqi Fan, Xiaojian Xue","doi":"10.1109/ICSAI57119.2022.10005357","DOIUrl":"https://doi.org/10.1109/ICSAI57119.2022.10005357","url":null,"abstract":"with the growing concerns of the current network security, development of hybrid cryptosystem based on the combination of optical symmetric cryptography and public key cryptography has aroused considerable research interest. In optical symmetry system, a new method for encrypting optical information using Fresnel lenses is proposed. The symmetry system’s keys include the focal length (f) and propagation wavelength ($lambda$). Then, in public key cryptography, the key ($lambda$,J) is encrypted by the RSA algorithm. Finally, we have implemented and result have been obtained using 12-level digital Fresnel lens by COMSOL Multiphysics to study of how the optical image deciphered the information using encryption keys. In addition, the main challenges and future prospects for the further development of hybrid cryptosystem system are discussed.","PeriodicalId":339547,"journal":{"name":"2022 8th International Conference on Systems and Informatics (ICSAI)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114176120","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}
Pub Date : 2022-12-10DOI: 10.1109/ICSAI57119.2022.10005521
Shuaichen Li, Jian-Feng Wu
Measuring the height of a site is of great significance for scientific research and daily life. This paper aims to measure the indoor and outdoor air pressure height, compare and analyze the error and fluctuation of indoor and outdoor measured data, and filter the measured value with Kalman filter to reduce the error and fluctuation; The influence of wind speed on measurement error and fluctuation is analyzed; Three methods are used to process the data of the wind measurement group and the results are compared; The predictability of air pressure sensor is tested and analyzed. The experimental results show that: 1) Compared with indoor measurement, the error and fluctuation of outdoor measurement are significantly worse. 2) In indoor and outdoor measurement, Kalman filter has a good effect, which can reduce the average error of measurement data by 31.9% at most and the measurement fluctuation by 37.5%. The effect of Kalman filtering is greatly affected by the measured data. The greater the fluctuation and error of the measured value, the more obvious the effect of Kalman filter. 3) Wind has a great influence on barometric altimetry, and the fluctuation and error of the measurement results increase with the increase of wind speed. Combine theoretical knowledge with the results of comparative experiments It is speculated that the reason for the large outdoor measurement error and fluctuation includes the interference of wind to the sensor. 4) The traditional least squares fitting is the worst, and the least squares fitting based on Kalman filter is the best. 5) In the windy state, the RMS of quadratic fitting is 60% higher than that in the windless state, and the residual error is 63.8% higher.
{"title":"Test and Analysis Of Indoor And Outdoor Barometric Altimetry","authors":"Shuaichen Li, Jian-Feng Wu","doi":"10.1109/ICSAI57119.2022.10005521","DOIUrl":"https://doi.org/10.1109/ICSAI57119.2022.10005521","url":null,"abstract":"Measuring the height of a site is of great significance for scientific research and daily life. This paper aims to measure the indoor and outdoor air pressure height, compare and analyze the error and fluctuation of indoor and outdoor measured data, and filter the measured value with Kalman filter to reduce the error and fluctuation; The influence of wind speed on measurement error and fluctuation is analyzed; Three methods are used to process the data of the wind measurement group and the results are compared; The predictability of air pressure sensor is tested and analyzed. The experimental results show that: 1) Compared with indoor measurement, the error and fluctuation of outdoor measurement are significantly worse. 2) In indoor and outdoor measurement, Kalman filter has a good effect, which can reduce the average error of measurement data by 31.9% at most and the measurement fluctuation by 37.5%. The effect of Kalman filtering is greatly affected by the measured data. The greater the fluctuation and error of the measured value, the more obvious the effect of Kalman filter. 3) Wind has a great influence on barometric altimetry, and the fluctuation and error of the measurement results increase with the increase of wind speed. Combine theoretical knowledge with the results of comparative experiments It is speculated that the reason for the large outdoor measurement error and fluctuation includes the interference of wind to the sensor. 4) The traditional least squares fitting is the worst, and the least squares fitting based on Kalman filter is the best. 5) In the windy state, the RMS of quadratic fitting is 60% higher than that in the windless state, and the residual error is 63.8% higher.","PeriodicalId":339547,"journal":{"name":"2022 8th International Conference on Systems and Informatics (ICSAI)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122106834","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}
Pub Date : 2022-12-10DOI: 10.1109/ICSAI57119.2022.10005466
H. Ma, Jianyin Fan, Qiang Wang
Deep reinforcement learning has been widely used in table tennis decision-making tasks, but most methods have their own defects, such as relying on high-precision trajectory prediction work or requiring targeted class training, etc. None of the methods can directly get a complete hitting strategy through the initial state of the ball. In this paper, we train a ping-pong hitting policy controller with model-free reinforcement learning. By extending the multi-dimensional Q-function, the prediction part of the table tennis task and the batting strategy part are integrated, the trajectory prediction work and the batting work are completed by a single network, which simplifies the complex prediction process and does not need to build a complex dynamics model or train a neural network to predict the trajectory of ping-pong balls. In this way, the deep reinforcement learning process and supervised trajectory prediction training process are organized into a single process. The experimental results show that the best convergence effect can be basically achieved in 50,000 rounds of training. 10,000 tests were performed as a test set with a success rate of over 99%.
{"title":"A Novel Ping-pong Task Strategy Based on Model-free Multi-dimensional Q-function Deep Reinforcement Learning","authors":"H. Ma, Jianyin Fan, Qiang Wang","doi":"10.1109/ICSAI57119.2022.10005466","DOIUrl":"https://doi.org/10.1109/ICSAI57119.2022.10005466","url":null,"abstract":"Deep reinforcement learning has been widely used in table tennis decision-making tasks, but most methods have their own defects, such as relying on high-precision trajectory prediction work or requiring targeted class training, etc. None of the methods can directly get a complete hitting strategy through the initial state of the ball. In this paper, we train a ping-pong hitting policy controller with model-free reinforcement learning. By extending the multi-dimensional Q-function, the prediction part of the table tennis task and the batting strategy part are integrated, the trajectory prediction work and the batting work are completed by a single network, which simplifies the complex prediction process and does not need to build a complex dynamics model or train a neural network to predict the trajectory of ping-pong balls. In this way, the deep reinforcement learning process and supervised trajectory prediction training process are organized into a single process. The experimental results show that the best convergence effect can be basically achieved in 50,000 rounds of training. 10,000 tests were performed as a test set with a success rate of over 99%.","PeriodicalId":339547,"journal":{"name":"2022 8th International Conference on Systems and Informatics (ICSAI)","volume":"52 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117084862","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}
Pub Date : 2022-12-10DOI: 10.1109/ICSAI57119.2022.10005340
Wenfeng Tian, Hui Deng, Pengfei Zeng
In order to better support the self-learning and dynamic upgrading of broadband RF pulse signal intelligent detection terminal for interference, partial discharge defective pulse signal features and its extraction algorithm, combined with the multi-sensor fusion, resource limitation and real-time information processing faced by the intelligent detection terminal, this paper is based on the principle of lightweight Docker container technology architecture, the partial discharge algorithm is developed, released, deployed, and The paper also adopts the multi-agent resource scheduling of lightweight containers to realize the intelligent scheduling of resources and continuous iteration of algorithms for broadband RF partial discharge detection system, supporting the development and application of broadband RF pulse signal detection and real-time analysis devices.
{"title":"Design of a Container-based Broadband RF Pulse Signal Analysis Algorithm Integration Scheme","authors":"Wenfeng Tian, Hui Deng, Pengfei Zeng","doi":"10.1109/ICSAI57119.2022.10005340","DOIUrl":"https://doi.org/10.1109/ICSAI57119.2022.10005340","url":null,"abstract":"In order to better support the self-learning and dynamic upgrading of broadband RF pulse signal intelligent detection terminal for interference, partial discharge defective pulse signal features and its extraction algorithm, combined with the multi-sensor fusion, resource limitation and real-time information processing faced by the intelligent detection terminal, this paper is based on the principle of lightweight Docker container technology architecture, the partial discharge algorithm is developed, released, deployed, and The paper also adopts the multi-agent resource scheduling of lightweight containers to realize the intelligent scheduling of resources and continuous iteration of algorithms for broadband RF partial discharge detection system, supporting the development and application of broadband RF pulse signal detection and real-time analysis devices.","PeriodicalId":339547,"journal":{"name":"2022 8th International Conference on Systems and Informatics (ICSAI)","volume":"24 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131037613","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}
This paper presents a bimodal emotion recognition system based on the voice and text information using small amount of speech data. Specifically, speech is divided into voice and text to learn emotion classifiers for each modal. The probabilities obtained from these emotion classifiers are weighted based on Mehrabian’s rule and summed up for each emotion to calculate the final score for bimodal emotion recognition. To create a highly accurate system while solving the problem that there are few Japanese speech data with emotion labels, we propose a novel data augmentation method and employ a transfer learning approach based on a pre-trained VGG16 model and a fine-tuned Bidirectional Encoder Representations from Transformers (BERT) model on the tweets. In order to prove the effectiveness of the proposed method, we revealed the recognition results for 7 emotional states, i.e., anger, sadness, joy, fear, surprise, disgust, and neutral. The experiment result suggested that our novel data augmentation method improved the accuracy and that bimodal predictions based on voice and text-based outperformed the single-model predictions.
{"title":"Learning a Bimodal Emotion Recognition System Based on Small Amount of Speech Data","authors":"Junya Furutani, Xin Kang, Keita Kiuchi, Ryota Nishimura, M. Sasayama, Kazuyuki Matsumoto","doi":"10.1109/ICSAI57119.2022.10005454","DOIUrl":"https://doi.org/10.1109/ICSAI57119.2022.10005454","url":null,"abstract":"This paper presents a bimodal emotion recognition system based on the voice and text information using small amount of speech data. Specifically, speech is divided into voice and text to learn emotion classifiers for each modal. The probabilities obtained from these emotion classifiers are weighted based on Mehrabian’s rule and summed up for each emotion to calculate the final score for bimodal emotion recognition. To create a highly accurate system while solving the problem that there are few Japanese speech data with emotion labels, we propose a novel data augmentation method and employ a transfer learning approach based on a pre-trained VGG16 model and a fine-tuned Bidirectional Encoder Representations from Transformers (BERT) model on the tweets. In order to prove the effectiveness of the proposed method, we revealed the recognition results for 7 emotional states, i.e., anger, sadness, joy, fear, surprise, disgust, and neutral. The experiment result suggested that our novel data augmentation method improved the accuracy and that bimodal predictions based on voice and text-based outperformed the single-model predictions.","PeriodicalId":339547,"journal":{"name":"2022 8th International Conference on Systems and Informatics (ICSAI)","volume":"105 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-12-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125541241","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：