Aiming at the problem of insufficient understanding of features and time-consuming calculation of long short-term memory networks in convolutional neural networks when extracting features, and the problem that both of them cannot reflect the importance of each word in the whole when extracting features, a method of law recommendation based on self-attention mechanism and feature fusion is proposed. Firstly, the text is preprocessed and Word2vec is used for word vectorization. Then the BIGRU model is used to extract the context features of the text, and the self-attention mechanism is added to extract the weighted information after the BIGRU features are extracted. CNN model is used to extract local features of text; finally, the characteristics of attention mechanism and CNN are fused to effectively solve the problems existing in a single model. The experimental results of the data set from the Judiciary Artificial Intelligence Challenge of China Law Research Cup show that the proposed model is better than the single model and its improved model.
{"title":"Law Recommendation Based on Self - Attention Mechanism and Feature Fusion","authors":"Lei Liu, Dezhi An","doi":"10.1145/3520084.3520101","DOIUrl":"https://doi.org/10.1145/3520084.3520101","url":null,"abstract":"Aiming at the problem of insufficient understanding of features and time-consuming calculation of long short-term memory networks in convolutional neural networks when extracting features, and the problem that both of them cannot reflect the importance of each word in the whole when extracting features, a method of law recommendation based on self-attention mechanism and feature fusion is proposed. Firstly, the text is preprocessed and Word2vec is used for word vectorization. Then the BIGRU model is used to extract the context features of the text, and the self-attention mechanism is added to extract the weighted information after the BIGRU features are extracted. CNN model is used to extract local features of text; finally, the characteristics of attention mechanism and CNN are fused to effectively solve the problems existing in a single model. The experimental results of the data set from the Judiciary Artificial Intelligence Challenge of China Law Research Cup show that the proposed model is better than the single model and its improved model.","PeriodicalId":444957,"journal":{"name":"Proceedings of the 2022 5th International Conference on Software Engineering and Information Management","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127776491","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}
K-means clustering is a widely used algorithm in cluster analysis. However, the selection of initial seeds determines the results of K-means clustering. The conventional K-means algorithm usually adopts the random strategy to select initial seeds, which is unable to generate an ideal clustering result in many cases. To solve the problem of the existing initial seeds selection (abbreviated to ISS) strategies for K-means clustering, we propose a novel initial seeds selection algorithm, called ISS_OD, based on outlier detection. In ISS_OD, we select the initial seeds of K-means clustering by calculating the distance outlier factor of every object, the weighted density of every object and the weighted distances between objects. Experimental results on several UCI datasets demonstrate the effectiveness of our algorithm for the ISS of K-means clustering.
k -均值聚类是聚类分析中应用广泛的一种算法。然而，初始种子的选择决定了K-means聚类的结果。传统的K-means算法通常采用随机策略选择初始种子，在很多情况下无法产生理想的聚类结果。为了解决现有K-means聚类初始种子选择(简称ISS)策略存在的问题，提出了一种基于离群点检测的初始种子选择算法ISS_OD。在ISS_OD中，我们通过计算每个目标的距离离群因子、每个目标的加权密度和目标之间的加权距离来选择K-means聚类的初始种子。在多个UCI数据集上的实验结果证明了该算法对K-means聚类的ISS的有效性。
{"title":"Initial Seeds Selection for K-means Clustering Based on Outlier Detection","authors":"Zhiyong Yang, Feng Jiang, J. Yu, Junwei Du","doi":"10.1145/3520084.3520106","DOIUrl":"https://doi.org/10.1145/3520084.3520106","url":null,"abstract":"K-means clustering is a widely used algorithm in cluster analysis. However, the selection of initial seeds determines the results of K-means clustering. The conventional K-means algorithm usually adopts the random strategy to select initial seeds, which is unable to generate an ideal clustering result in many cases. To solve the problem of the existing initial seeds selection (abbreviated to ISS) strategies for K-means clustering, we propose a novel initial seeds selection algorithm, called ISS_OD, based on outlier detection. In ISS_OD, we select the initial seeds of K-means clustering by calculating the distance outlier factor of every object, the weighted density of every object and the weighted distances between objects. Experimental results on several UCI datasets demonstrate the effectiveness of our algorithm for the ISS of K-means clustering.","PeriodicalId":444957,"journal":{"name":"Proceedings of the 2022 5th International Conference on Software Engineering and Information Management","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127889893","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}
With the development of big data technology, data accessed by big data platforms maintain the features of mass, isomerism, heterogeneous, and streaming. Therefore, how to access the varied data sources of isomerism and heterogeneous data and how to process and analyze the data become the current challenges. In this paper, we design and implement a data aggregation and anomaly detection system for isomerism and heterogeneous data. The system proposes a novel isomerism and heterogeneous data access sub-system. The sub-system applies improved Avro as the unified data description format and presents different storage algorithms for data serialization to raise the data adaption efficiency. The system adopts Kafka as the message middleware for data aggregation and distribution. Also, we design the anomaly detection and alarming sub-system for detecting the anomalies of streaming data on time and notifying the users. The data aggregation and anomaly detection system has passed all the tests and applied in small and medium-sized enterprises.
{"title":"Data Aggregation and Anomaly Detection System for Isomerism and Heterogeneous Data","authors":"Yunze Li, Yuxuan Wu, Ruisen Tang","doi":"10.1145/3520084.3520099","DOIUrl":"https://doi.org/10.1145/3520084.3520099","url":null,"abstract":"With the development of big data technology, data accessed by big data platforms maintain the features of mass, isomerism, heterogeneous, and streaming. Therefore, how to access the varied data sources of isomerism and heterogeneous data and how to process and analyze the data become the current challenges. In this paper, we design and implement a data aggregation and anomaly detection system for isomerism and heterogeneous data. The system proposes a novel isomerism and heterogeneous data access sub-system. The sub-system applies improved Avro as the unified data description format and presents different storage algorithms for data serialization to raise the data adaption efficiency. The system adopts Kafka as the message middleware for data aggregation and distribution. Also, we design the anomaly detection and alarming sub-system for detecting the anomalies of streaming data on time and notifying the users. The data aggregation and anomaly detection system has passed all the tests and applied in small and medium-sized enterprises.","PeriodicalId":444957,"journal":{"name":"Proceedings of the 2022 5th International Conference on Software Engineering and Information Management","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126235699","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 proliferation of outlets for news media in recent decades has contributed to faster issuance of news data. News analysis has been one of the key activities conducted by researchers in a broad variety of research disciplines. In general, the analysis process used in these studies includes interpreting the content of the news items, and then discovering their impact in a specific area. In this paper, we delve into the field of news analysis applied to the financial domain and explore news sentiment impact analysis in the context of financial markets. Existing studies lack systematic methods to assimilate financial context and evaluate the impact of a given news dataset on relevant entities financial market performance. We introduce an improved version of the framework called News Sentiment Impact Analysis (NSIA) that encompasses models, supporting software architecture and processes for defining various financial contexts and conducting news sentiment impact analysis. The framework is then evaluated using a prototype implementation and a case study that investigates the impact of extremely negative news on the stock price of the related entities. The results demonstrate the functionality, usability and reproducibility of the framework, and its capability to bridge the gap between generating news sentiment and evaluating its impact in selected financial contexts.
{"title":"A Framework for Facilitating Reproducible News Sentiment Impact Analysis","authors":"Weisi Chen, Islam Al-Qudah, F. Rabhi","doi":"10.1145/3520084.3520104","DOIUrl":"https://doi.org/10.1145/3520084.3520104","url":null,"abstract":"The proliferation of outlets for news media in recent decades has contributed to faster issuance of news data. News analysis has been one of the key activities conducted by researchers in a broad variety of research disciplines. In general, the analysis process used in these studies includes interpreting the content of the news items, and then discovering their impact in a specific area. In this paper, we delve into the field of news analysis applied to the financial domain and explore news sentiment impact analysis in the context of financial markets. Existing studies lack systematic methods to assimilate financial context and evaluate the impact of a given news dataset on relevant entities financial market performance. We introduce an improved version of the framework called News Sentiment Impact Analysis (NSIA) that encompasses models, supporting software architecture and processes for defining various financial contexts and conducting news sentiment impact analysis. The framework is then evaluated using a prototype implementation and a case study that investigates the impact of extremely negative news on the stock price of the related entities. The results demonstrate the functionality, usability and reproducibility of the framework, and its capability to bridge the gap between generating news sentiment and evaluating its impact in selected financial contexts.","PeriodicalId":444957,"journal":{"name":"Proceedings of the 2022 5th International Conference on Software Engineering and Information Management","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122716116","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}
∗ Under the wave of digital transformation of traditional enterprises, large numbers of enterprises are actively embracing cloud platforms, leading to the gradual increase in the operation and maintenance cost of the information system. Furthermore, cloud platform health status assessment is increasingly important for enterprises, while operation and maintenance methods of the traditional cloud platform would increase manpower and material resources costs of the enterprises. This paper proposes to design a cloud platform health status assessment application that supports the continuous evolution of assessment capabilities (HSAACE). The HSAACE has the openness of assessment applications, the evolution of assessment capabilities, and the integrity of health assessment. Moreover, it can not only adapt to the changes between different platforms but also effectively deal with the objective characteristics of the data stream distribution and structure, which are constantly changing over time, thereby potentially reducing the operation and maintenance cost of the enterprise system.
{"title":"HSAACE: Design a Cloud Platform Health Status Assessment Application to Support Continuous Evolution of Assessment Capabilities","authors":"Yu-Shu Hu, Yunxuan Wang","doi":"10.1145/3520084.3520105","DOIUrl":"https://doi.org/10.1145/3520084.3520105","url":null,"abstract":"∗ Under the wave of digital transformation of traditional enterprises, large numbers of enterprises are actively embracing cloud platforms, leading to the gradual increase in the operation and maintenance cost of the information system. Furthermore, cloud platform health status assessment is increasingly important for enterprises, while operation and maintenance methods of the traditional cloud platform would increase manpower and material resources costs of the enterprises. This paper proposes to design a cloud platform health status assessment application that supports the continuous evolution of assessment capabilities (HSAACE). The HSAACE has the openness of assessment applications, the evolution of assessment capabilities, and the integrity of health assessment. Moreover, it can not only adapt to the changes between different platforms but also effectively deal with the objective characteristics of the data stream distribution and structure, which are constantly changing over time, thereby potentially reducing the operation and maintenance cost of the enterprise system.","PeriodicalId":444957,"journal":{"name":"Proceedings of the 2022 5th International Conference on Software Engineering and Information Management","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114405528","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}
P. Netinant, Auttapon Amatyakul, Meennapa Rukhiran
The Internet of Things (IoT) is a new paradigm that connects the Internet and physical objects across a range of industries, including home automation, industrial processes, human health, and environmental monitoring. It raises the significance of Internet-connected devices in our daily lives, bringing a flood of benefits and concerns about security. For decades, home intruder detection systems have been an integral part of home security systems. However, implementing intruder motion detection techniques is difficult due to the IoT's unique characteristics, such as resource-constrained devices. In this study, we offer a prototype for an Internet of Things-based home intruder detection system. Our objective is to discover areas for development and practice, as well as research opportunities and concerns. Additionally, we examined several choices for each attribute, including aspects of passive infrared motion detector works that suggest unique methods for home invader motion detectors on the Internet of Things. Detail aspects of passive infrared motion detector work range a home intruder motion-detecting schemes. The primary evaluation of the passive infrared motion detector system was conducted to evaluate functional detection. The system can do daily motion detection work automatically—it provides intruder detection in a variety of distances and angles of circumstances and location monitoring.
{"title":"Alert Intruder Detection System Using Passive Infrared Motion Detector based on Internet of Things","authors":"P. Netinant, Auttapon Amatyakul, Meennapa Rukhiran","doi":"10.1145/3520084.3520112","DOIUrl":"https://doi.org/10.1145/3520084.3520112","url":null,"abstract":"The Internet of Things (IoT) is a new paradigm that connects the Internet and physical objects across a range of industries, including home automation, industrial processes, human health, and environmental monitoring. It raises the significance of Internet-connected devices in our daily lives, bringing a flood of benefits and concerns about security. For decades, home intruder detection systems have been an integral part of home security systems. However, implementing intruder motion detection techniques is difficult due to the IoT's unique characteristics, such as resource-constrained devices. In this study, we offer a prototype for an Internet of Things-based home intruder detection system. Our objective is to discover areas for development and practice, as well as research opportunities and concerns. Additionally, we examined several choices for each attribute, including aspects of passive infrared motion detector works that suggest unique methods for home invader motion detectors on the Internet of Things. Detail aspects of passive infrared motion detector work range a home intruder motion-detecting schemes. The primary evaluation of the passive infrared motion detector system was conducted to evaluate functional detection. The system can do daily motion detection work automatically—it provides intruder detection in a variety of distances and angles of circumstances and location monitoring.","PeriodicalId":444957,"journal":{"name":"Proceedings of the 2022 5th International Conference on Software Engineering and Information Management","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121641773","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}
Testing the non-functional requirements (NFR) of a system is particularly complicated and time-consuming. Challenges in this area are compounded when the system is developed under some offspring of Agile methodologies, which favor iterative development and rapid feedback from extensive testing. The authors of this paper build upon previous work investigating the common challenges and solutions cited in recent peer-reviewed research on this topic to design and build a tool consolidating many of the concepts found in this investigation. The tool is known as LuluPerfTest (LPT) and is an NFR testing framework meant to plug into continuous integration (CI) systems to run NFR tests configured with a JSON script. This allows developers and testers to build maintainable and minimally complex automated NFR test scripts. This study explains the challenges inherent in NFR testing in Agile software development and presents how LPT confronts those challenges. It aims to explain LPT and invite collaboration among other testing, verification, and validation researchers to create an open sources software (OSS) solution to the problems of NFR testing in Agile software development projects.
{"title":"Design and Development of a Technology-Agnostic NFR Testing Framework: Introducing the framework and discussing the future of load testing in Agile software development","authors":"Erik Whiting, Soma Datta","doi":"10.1145/3520084.3520092","DOIUrl":"https://doi.org/10.1145/3520084.3520092","url":null,"abstract":"Testing the non-functional requirements (NFR) of a system is particularly complicated and time-consuming. Challenges in this area are compounded when the system is developed under some offspring of Agile methodologies, which favor iterative development and rapid feedback from extensive testing. The authors of this paper build upon previous work investigating the common challenges and solutions cited in recent peer-reviewed research on this topic to design and build a tool consolidating many of the concepts found in this investigation. The tool is known as LuluPerfTest (LPT) and is an NFR testing framework meant to plug into continuous integration (CI) systems to run NFR tests configured with a JSON script. This allows developers and testers to build maintainable and minimally complex automated NFR test scripts. This study explains the challenges inherent in NFR testing in Agile software development and presents how LPT confronts those challenges. It aims to explain LPT and invite collaboration among other testing, verification, and validation researchers to create an open sources software (OSS) solution to the problems of NFR testing in Agile software development projects.","PeriodicalId":444957,"journal":{"name":"Proceedings of the 2022 5th International Conference on Software Engineering and Information Management","volume":"120 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121823531","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}
In order to achieve the goals of carbon peak and carbon neutralization, the reform and innovation of green finance has also been put on the agenda. Commercial banks as an important part of the national economy, the innovation of green credit model can not only promote the development of green finance, but also lead the high-quality development of China's environmental protection industry. However, because the green credit model of China's commercial banks started late and developed slowly, there are still some problems in many aspects, which need to be further optimized and adjusted. Based on this, this paper first expounds the existing main green credit models of China's commercial banks. Secondly, taking China Construction Bank as an example, it focuses on the green credit model of the pilot Bank of China Construction Bank. It is found that the overall proportion of green credit model is relatively low, the green credit model is single and lack for innovation as well as specialization. Finally, aiming at the above problems, this paper puts forward specific suggestions to promote the innovation and development of green credit model of commercial banks in China.
{"title":"Research on the Innovation of Commercial Banks' Green Finance Credit Model: Based on the Case of China Construction Bank","authors":"Liu Yang, Sheng-Hung Su, Quanxin Gan","doi":"10.1145/3520084.3520121","DOIUrl":"https://doi.org/10.1145/3520084.3520121","url":null,"abstract":"In order to achieve the goals of carbon peak and carbon neutralization, the reform and innovation of green finance has also been put on the agenda. Commercial banks as an important part of the national economy, the innovation of green credit model can not only promote the development of green finance, but also lead the high-quality development of China's environmental protection industry. However, because the green credit model of China's commercial banks started late and developed slowly, there are still some problems in many aspects, which need to be further optimized and adjusted. Based on this, this paper first expounds the existing main green credit models of China's commercial banks. Secondly, taking China Construction Bank as an example, it focuses on the green credit model of the pilot Bank of China Construction Bank. It is found that the overall proportion of green credit model is relatively low, the green credit model is single and lack for innovation as well as specialization. Finally, aiming at the above problems, this paper puts forward specific suggestions to promote the innovation and development of green credit model of commercial banks in China.","PeriodicalId":444957,"journal":{"name":"Proceedings of the 2022 5th International Conference on Software Engineering and Information Management","volume":"73 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124692688","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}
P. Netinant, Krairat Arpabusayapan, Meennapa Rukhiran
The Internet of Things has been substantially developed for disabled and elderly persons in various domains. Speech recognition is an extremely challenging technique for cost-effective human contact, communication, and control. Numerous experiments have been undertaken on voice recognition systems in order to provide a more complete explanation of language commands, particularly for non-native English speakers and languages with tone variations. This article outlines the development of a Raspberry Pi-based spoken command system. The system was developed and installed using Python, and it makes use of the Google Speech Recognition API as a speech-to-text converter. Our light control system's speech recognition system is capable of receiving voice commands via a USB microphone. The experimental results compare the accuracy of light control for Thai and English orders utilizing individuals who are Thai elderly speakers. Thai speech is recognized more precisely than English speech by the suggested approach. These startling findings refute the concept that speech recognition algorithms can boost the growth of the Internet of Things. However, the system's accuracy in recognizing speech for disabled and elderly users should be weighed against the country's national or indigenous languages.
{"title":"Speech Recognition for Light Control on Raspberry Pi Using Python Programming","authors":"P. Netinant, Krairat Arpabusayapan, Meennapa Rukhiran","doi":"10.1145/3520084.3520090","DOIUrl":"https://doi.org/10.1145/3520084.3520090","url":null,"abstract":"The Internet of Things has been substantially developed for disabled and elderly persons in various domains. Speech recognition is an extremely challenging technique for cost-effective human contact, communication, and control. Numerous experiments have been undertaken on voice recognition systems in order to provide a more complete explanation of language commands, particularly for non-native English speakers and languages with tone variations. This article outlines the development of a Raspberry Pi-based spoken command system. The system was developed and installed using Python, and it makes use of the Google Speech Recognition API as a speech-to-text converter. Our light control system's speech recognition system is capable of receiving voice commands via a USB microphone. The experimental results compare the accuracy of light control for Thai and English orders utilizing individuals who are Thai elderly speakers. Thai speech is recognized more precisely than English speech by the suggested approach. These startling findings refute the concept that speech recognition algorithms can boost the growth of the Internet of Things. However, the system's accuracy in recognizing speech for disabled and elderly users should be weighed against the country's national or indigenous languages.","PeriodicalId":444957,"journal":{"name":"Proceedings of the 2022 5th International Conference on Software Engineering and Information Management","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131062088","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}
Although the researchers have made great progress in the field of text detection and text recognition, text detection and text recognition are still facing great challenges because of the differences of text fonts and the complexity of backgrounds. Traditional text detection and recognition methods rely on artificial designed features and rules, thus the methods usually requires higher text layout and text resolution. Aiming at the problem of automatic book inventory in library, the paper proposes a new method based on an EAST (Efficient and Accurate Scene Text Detector) detection and an CRNN (Continuous Recurrent Neural Network) recognition. In this method, the library book titles are detected by EAST to get the text area on the side of the books and also to output coordinates. Then, the content of the text area is further identified by the CRNN. Finally, through the comparison of the database, we know whether books of libraries are on corresponding shelves or not. The experimental results show that this method can quickly and accurately realize the task of automatic book title recognitions, and it can still effectively detect the text area and accurately recognize the book title in the case of dark light. Therefore, the method effectively solves the problem of manual inventory of books in existing libraries, which is time-consuming and laborious, and has a certain engineering application prospect.
{"title":"Automatic inventory system of librarian books based on a deep learning algorithm with EAST and CRNN","authors":"Shuanle Wang, Chaoyi Dong, Peng Yang, Chen Xiaoyan, Zang Weidong","doi":"10.1145/3520084.3520119","DOIUrl":"https://doi.org/10.1145/3520084.3520119","url":null,"abstract":"Although the researchers have made great progress in the field of text detection and text recognition, text detection and text recognition are still facing great challenges because of the differences of text fonts and the complexity of backgrounds. Traditional text detection and recognition methods rely on artificial designed features and rules, thus the methods usually requires higher text layout and text resolution. Aiming at the problem of automatic book inventory in library, the paper proposes a new method based on an EAST (Efficient and Accurate Scene Text Detector) detection and an CRNN (Continuous Recurrent Neural Network) recognition. In this method, the library book titles are detected by EAST to get the text area on the side of the books and also to output coordinates. Then, the content of the text area is further identified by the CRNN. Finally, through the comparison of the database, we know whether books of libraries are on corresponding shelves or not. The experimental results show that this method can quickly and accurately realize the task of automatic book title recognitions, and it can still effectively detect the text area and accurately recognize the book title in the case of dark light. Therefore, the method effectively solves the problem of manual inventory of books in existing libraries, which is time-consuming and laborious, and has a certain engineering application prospect.","PeriodicalId":444957,"journal":{"name":"Proceedings of the 2022 5th International Conference on Software Engineering and Information Management","volume":"35 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-01-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133755143","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：