One of the most challenging problems faced by ecologists and other biological re- searchers today is to analyze the massive amounts of data being collected by advanced monitoring systems like camera traps, wireless sensor networks, high-frequency radio track- ers, global positioning systems, and satellite tracking systems being used today. It has become expensive, laborious, and time-consuming to analyze this huge data using man- ual and traditional statistical techniques. Recent developments in the deep learning field are showing promising results towards automating the analysis of these extremely large datasets. The primary objective of this study was to test the capabilities of the state-of- the-art deep learning architectures to detect birds in the webcam captured images. A total of 10592 images were collected for this study from the Cornell Lab of Ornithology live stream feeds situated in six unique locations in United States, Ecuador, New Zealand, and Panama. To achieve the main objective of the study, we studied and evaluated two con- volutional neural network object detection meta-architectures, single-shot detector (SSD) and Faster R-CNN in combination with MobileNet-V2, ResNet50, ResNet101, ResNet152, and Inception ResNet-V2 feature extractors. Through transfer learning, all the models were initialized using weights pre-trained on the MS COCO (Microsoft Common Objects in Context) dataset provided by TensorFlow 2 object detection API. The Faster R-CNN model coupled with ResNet152 outperformed all other models with a mean average preci- sion of 92.3%. However, the SSD model with the MobileNet-V2 feature extraction network achieved the lowest inference time (110ms) and the smallest memory capacity (30.5MB) compared to its counterparts. The outstanding results achieved in this study confirm that deep learning-based algorithms are capable of detecting birds of different sizes in differ- ent environments and the best model could potentially help ecologists in monitoring and identifying birds from other species.
当今生态学家和其他生物研究人员面临的最具挑战性的问题之一是分析由先进的监测系统收集的大量数据，这些系统包括摄像机陷阱、无线传感器网络、高频无线电跟踪器、全球定位系统和卫星跟踪系统。使用人工和传统的统计技术来分析这些庞大的数据已经变得昂贵、费力和耗时。深度学习领域的最新发展显示了对这些超大数据集的自动化分析的有希望的结果。本研究的主要目的是测试最先进的深度学习架构在网络摄像头捕获的图像中检测鸟类的能力。本研究共收集了10592张图像，这些图像来自康奈尔鸟类学实验室位于美国、厄瓜多尔、新西兰和巴拿马六个独特地点的直播饲料。为了实现研究的主要目标，我们研究并评估了两种卷积神经网络目标检测元架构，即单次检测(SSD)和Faster R-CNN，结合MobileNet-V2、ResNet50、ResNet101、ResNet152和Inception ResNet-V2特征提取器。通过迁移学习，使用TensorFlow 2对象检测API提供的MS COCO (Microsoft Common Objects in Context)数据集预训练的权重对所有模型进行初始化。与ResNet152结合的Faster R-CNN模型以92.3%的平均精度优于所有其他模型。然而，与同类模型相比，具有MobileNet-V2特征提取网络的SSD模型实现了最低的推理时间(110ms)和最小的内存容量(30.5MB)。本研究取得的突出结果证实，基于深度学习的算法能够在不同环境中检测不同大小的鸟类，最佳模型可能有助于生态学家监测和识别其他物种的鸟类。
{"title":"Automating Bird Detection Based on Webcam Captured Images using Deep Learning","authors":"Alex Mirugwe, Juwa Nyirenda, Emmanuel Dufourq","doi":"10.29007/9fr5","DOIUrl":"https://doi.org/10.29007/9fr5","url":null,"abstract":"One of the most challenging problems faced by ecologists and other biological re- searchers today is to analyze the massive amounts of data being collected by advanced monitoring systems like camera traps, wireless sensor networks, high-frequency radio track- ers, global positioning systems, and satellite tracking systems being used today. It has become expensive, laborious, and time-consuming to analyze this huge data using man- ual and traditional statistical techniques. Recent developments in the deep learning field are showing promising results towards automating the analysis of these extremely large datasets. The primary objective of this study was to test the capabilities of the state-of- the-art deep learning architectures to detect birds in the webcam captured images. A total of 10592 images were collected for this study from the Cornell Lab of Ornithology live stream feeds situated in six unique locations in United States, Ecuador, New Zealand, and Panama. To achieve the main objective of the study, we studied and evaluated two con- volutional neural network object detection meta-architectures, single-shot detector (SSD) and Faster R-CNN in combination with MobileNet-V2, ResNet50, ResNet101, ResNet152, and Inception ResNet-V2 feature extractors. Through transfer learning, all the models were initialized using weights pre-trained on the MS COCO (Microsoft Common Objects in Context) dataset provided by TensorFlow 2 object detection API. The Faster R-CNN model coupled with ResNet152 outperformed all other models with a mean average preci- sion of 92.3%. However, the SSD model with the MobileNet-V2 feature extraction network achieved the lowest inference time (110ms) and the smallest memory capacity (30.5MB) compared to its counterparts. The outstanding results achieved in this study confirm that deep learning-based algorithms are capable of detecting birds of different sizes in differ- ent environments and the best model could potentially help ecologists in monitoring and identifying birds from other species.","PeriodicalId":93549,"journal":{"name":"EPiC series in computing","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69427803","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}
Joshua Dahl, Erik Marsh, Christopher Lewis, Frederick Harris
Due to the rapidly evolving nature of the Virtual Reality field, many frameworks for multiuser interaction have become outdated, with few (if any) designed to support mixed virtual and non-virtual interactions. We have developed a framework that lays an exten- sible and forward-looking foundation for mixed interactions based upon a novel method of ensuring that inputs, visuals, and networking can all communicate without needing to understand the others’ internals. We tested this framework in the development of several applications and proved that it can easily be adapted to support application requirements it was not originally designed for.
{"title":"MuVR: A Multiuser Virtual Reality Framework for Unity","authors":"Joshua Dahl, Erik Marsh, Christopher Lewis, Frederick Harris","doi":"10.29007/jdlg","DOIUrl":"https://doi.org/10.29007/jdlg","url":null,"abstract":"Due to the rapidly evolving nature of the Virtual Reality field, many frameworks for multiuser interaction have become outdated, with few (if any) designed to support mixed virtual and non-virtual interactions. We have developed a framework that lays an exten- sible and forward-looking foundation for mixed interactions based upon a novel method of ensuring that inputs, visuals, and networking can all communicate without needing to understand the others’ internals. We tested this framework in the development of several applications and proved that it can easily be adapted to support application requirements it was not originally designed for.","PeriodicalId":93549,"journal":{"name":"EPiC series in computing","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69432538","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}
TheThe driver safety is given an utmost importance in the Transportation system. Road safety is mostly dependent on the all driver’s on road and their behavior. Aggressive driving behavior such as speed, braking, accelerations etc are some of the major factors contributing to the safety which can jeopardize human lives if a fatality occurs. To improve the safety of drivers and other road users, we proposed a framework which ranks and re- wards the driver’s behavior for each day in crypto tokens. Existing frameworks emphasizes on analyzing or ranking the behavior, however monetizing driver’s behavior will improve the driver’s discipline. A randomized simulated traffic is used to extract the both friendly and aggressive driving patterns and provide test crypto tokens based on them.
{"title":"DSRBT - Driving Safety Reward based on Blockchain Technology","authors":"Sruthi Rachamalla, H. Hexmoor","doi":"10.29007/pb3c","DOIUrl":"https://doi.org/10.29007/pb3c","url":null,"abstract":"TheThe driver safety is given an utmost importance in the Transportation system. Road safety is mostly dependent on the all driver’s on road and their behavior. Aggressive driving behavior such as speed, braking, accelerations etc are some of the major factors contributing to the safety which can jeopardize human lives if a fatality occurs. To improve the safety of drivers and other road users, we proposed a framework which ranks and re- wards the driver’s behavior for each day in crypto tokens. Existing frameworks emphasizes on analyzing or ranking the behavior, however monetizing driver’s behavior will improve the driver’s discipline. A randomized simulated traffic is used to extract the both friendly and aggressive driving patterns and provide test crypto tokens based on them.","PeriodicalId":93549,"journal":{"name":"EPiC series in computing","volume":"7 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69448013","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 the field of software development, the processes, technologies, and practices have matured over the time to achieve a higher level of delivery and quality. However, the de- velopment phase, which is an essential part of the software development life cycle (SDLC), is still consuming a significant cost (time and resources) in both approaches, waterfall and agile. The reason behind that, current technologies and approaches of software develop- ment are somehow following the same rules and practices for decades, and have not evolved with the proper velocity over the time. In this article, and based on real-life case studies, we will discuss how the utilization of components re-usability (API’s and frameworks), metadata-driven development, code generation, and Artificial Intelligence can make the software development more efficient by creating a holistic approach to creating software systems.
{"title":"Software Development: Past, Present, and Future","authors":"Jalal Kiswani, S. Dascalu, Fred Harris","doi":"10.29007/qzrd","DOIUrl":"https://doi.org/10.29007/qzrd","url":null,"abstract":"In the field of software development, the processes, technologies, and practices have matured over the time to achieve a higher level of delivery and quality. However, the de- velopment phase, which is an essential part of the software development life cycle (SDLC), is still consuming a significant cost (time and resources) in both approaches, waterfall and agile. The reason behind that, current technologies and approaches of software develop- ment are somehow following the same rules and practices for decades, and have not evolved with the proper velocity over the time. In this article, and based on real-life case studies, we will discuss how the utilization of components re-usability (API’s and frameworks), metadata-driven development, code generation, and Artificial Intelligence can make the software development more efficient by creating a holistic approach to creating software systems.","PeriodicalId":93549,"journal":{"name":"EPiC series in computing","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69450131","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}
Shunsuke Yokoyama, T. Miyahara, Yusuke Suzuki, Tomoyuki Uchida, T. Kuboyama
Machine learning and data mining from tree structured data are studied intensively. In this paper, as tree structured patterns we use tag tree patterns with vertex and edge labels and wildcards in order to represent label connecting relation of vertices and edges in tree structured data. We propose an evolutionary learning method based on Genetic Programming for acquiring characteristic tag tree patterns with vertex and edge labels and wildcards from positive and negative tree structured data. By using label information, that is, label connecting relation of positive examples, as inappropriate individuals we can exclude tag tree patterns that do not satisfy label connecting relation of positive examples. We report experimental results on our evolutionary learning method and show the effectiveness of using label connecting relation of positive examples.
{"title":"Using Label Information in a Genetic Programming Based Method for Acquiring Tag Tree Patterns with Vertex Labels and Wildcards","authors":"Shunsuke Yokoyama, T. Miyahara, Yusuke Suzuki, Tomoyuki Uchida, T. Kuboyama","doi":"10.29007/tfgn","DOIUrl":"https://doi.org/10.29007/tfgn","url":null,"abstract":"Machine learning and data mining from tree structured data are studied intensively. In this paper, as tree structured patterns we use tag tree patterns with vertex and edge labels and wildcards in order to represent label connecting relation of vertices and edges in tree structured data. We propose an evolutionary learning method based on Genetic Programming for acquiring characteristic tag tree patterns with vertex and edge labels and wildcards from positive and negative tree structured data. By using label information, that is, label connecting relation of positive examples, as inappropriate individuals we can exclude tag tree patterns that do not satisfy label connecting relation of positive examples. We report experimental results on our evolutionary learning method and show the effectiveness of using label connecting relation of positive examples.","PeriodicalId":93549,"journal":{"name":"EPiC series in computing","volume":"17 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69451650","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}
Valérie Le Strat, Patrik Maltusch, Esa Suominen, Lluís Alfons Ariño Martín
There are important lessons to be learnt from actual implementations of enterprise architecture and capability models in higher education. In this paper we draw on three different case studies from France, Finland, and Spain respectively, showcasing both commonalities and important differences. The examples showcase use cases as well as the organisations and processes behind the developmentsWe argue that one important contribution from these European examples is an understanding of the national differences that need to be accommodated in a standard such as the recently introduced higher education reference model (HERM). One aspect that also becomes obvious from a European perspective is the need for translations–and how language use is closely connected to local variations in the Higher Education models.
{"title":"Using enterprise architecture and capability models in higher education: case studies from the EUNIS community","authors":"Valérie Le Strat, Patrik Maltusch, Esa Suominen, Lluís Alfons Ariño Martín","doi":"10.29007/hxf1","DOIUrl":"https://doi.org/10.29007/hxf1","url":null,"abstract":"There are important lessons to be learnt from actual implementations of enterprise architecture and capability models in higher education. In this paper we draw on three different case studies from France, Finland, and Spain respectively, showcasing both commonalities and important differences. The examples showcase use cases as well as the organisations and processes behind the developmentsWe argue that one important contribution from these European examples is an understanding of the national differences that need to be accommodated in a standard such as the recently introduced higher education reference model (HERM). One aspect that also becomes obvious from a European perspective is the need for translations–and how language use is closely connected to local variations in the Higher Education models.","PeriodicalId":93549,"journal":{"name":"EPiC series in computing","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69432527","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}
Ian Macedo Maiwald Santos, Luciana de Oliveira Rech, Ricardo Moraes
Currently, many short texts are published online, especially on social media platforms. High impact events, for example, are highly commented on by users. Understanding the subjects and patterns hidden in online discussions is a very important task for contexts such as elections, natural disasters or major sporting events. However, many works of this nature use techniques that, despite showing satisfactory results, are not the most suitable when it comes to the short texts on social media and may suffer a loss in their results. Therefore, this paper presents a text mining method for messages published on social media, with a data pre-processing step and topic modeling for short texts. For this paper, we created a data set from real world tweets related to COVID-19 that is openly available1 for research purposes.
{"title":"A Topic Modeling Method for Analyzes of Short-Text Data in Social Media Networks","authors":"Ian Macedo Maiwald Santos, Luciana de Oliveira Rech, Ricardo Moraes","doi":"10.29007/kr1z","DOIUrl":"https://doi.org/10.29007/kr1z","url":null,"abstract":"Currently, many short texts are published online, especially on social media platforms. High impact events, for example, are highly commented on by users. Understanding the subjects and patterns hidden in online discussions is a very important task for contexts such as elections, natural disasters or major sporting events. However, many works of this nature use techniques that, despite showing satisfactory results, are not the most suitable when it comes to the short texts on social media and may suffer a loss in their results. Therefore, this paper presents a text mining method for messages published on social media, with a data pre-processing step and topic modeling for short texts. For this paper, we created a data set from real world tweets related to COVID-19 that is openly available1 for research purposes.","PeriodicalId":93549,"journal":{"name":"EPiC series in computing","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69433973","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 discussed how to build deep reinforcement learning (DRL) agents to determine the allocation of money for assets in a portfolio so that the maximum return can be gained. The policy gradient method from reinforcement learning and convolutional neural network/recurrent neural network/convolutional neural network concatenated with the recurrent neural network from deep learning are combined together to build the agents. With the proposed models, three types of portfolios are tested: stocks portfolio which has a positive influence due to the Covid-19, stocks portfolio which has a negative influence due to the Covid-19, and portfolio of stocks combined with cryptocurrency which are randomly selected. The performance of our DRL agents was compared with that of equal-weighted agent and all the money fully invested on one stock agents. All of our DRL agents showed the best performance on the randomly selected portfolio, which has an overall stable up-ticking trend. In addition, the performance of linear regression model was also tested with the random selected portfolio, and it shows a poor result compared to other agents.
{"title":"Deep Reinforcement Learning for Portfolio Management","authors":"Yue Ma, Ziping Liu, Chuck McAllister","doi":"10.29007/w2m3","DOIUrl":"https://doi.org/10.29007/w2m3","url":null,"abstract":"This paper discussed how to build deep reinforcement learning (DRL) agents to determine the allocation of money for assets in a portfolio so that the maximum return can be gained. The policy gradient method from reinforcement learning and convolutional neural network/recurrent neural network/convolutional neural network concatenated with the recurrent neural network from deep learning are combined together to build the agents. With the proposed models, three types of portfolios are tested: stocks portfolio which has a positive influence due to the Covid-19, stocks portfolio which has a negative influence due to the Covid-19, and portfolio of stocks combined with cryptocurrency which are randomly selected. The performance of our DRL agents was compared with that of equal-weighted agent and all the money fully invested on one stock agents. All of our DRL agents showed the best performance on the randomly selected portfolio, which has an overall stable up-ticking trend. In addition, the performance of linear regression model was also tested with the random selected portfolio, and it shows a poor result compared to other agents.","PeriodicalId":93549,"journal":{"name":"EPiC series in computing","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69452357","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 describes an automated grading system for MS-Excel files and MS-Word files for information technology education. The system can relieve teachers’ workloads to grade many exercises of MS-Excel/MS-Word files. It can also provide immediate feedback and has a mechanism to prevent students from submitting copied files.In addition, we discuss the system’s effectiveness from both perspectives: the time to grade MS-Excel/MS-Word files and the average normalized gain computed by the operation records of the system in our university.
{"title":"Implementation of an Automated Grading System for Microsoft Excel Spreadsheets and Word Documents","authors":"Kazunori Iwata, Yoshimitsu Matsui","doi":"10.29007/1zs6","DOIUrl":"https://doi.org/10.29007/1zs6","url":null,"abstract":"This paper describes an automated grading system for MS-Excel files and MS-Word files for information technology education. The system can relieve teachers’ workloads to grade many exercises of MS-Excel/MS-Word files. It can also provide immediate feedback and has a mechanism to prevent students from submitting copied files.In addition, we discuss the system’s effectiveness from both perspectives: the time to grade MS-Excel/MS-Word files and the average normalized gain computed by the operation records of the system in our university.","PeriodicalId":93549,"journal":{"name":"EPiC series in computing","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69420270","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}
Enterprise Architecture Management (EAM) is widely used in the public sector and is increasingly understood as a driver of digital transformation. After reviewing the current literature on the EAM Services in the public sector, this article reports on experiences in the realignment of the EAM services in Deutsche Rentenversicherung (DRV - German pension insurance), presents an approach supported by Value Proposition Canvas (VPC), and details the EAM services that were designed using the method.
{"title":"Designing Enterprise Architecture Management Services – A Transformation Journey in the Public Sector","authors":"H. Koç, Wilhelm Weisweber, Marcus Luettke","doi":"10.29007/7tfm","DOIUrl":"https://doi.org/10.29007/7tfm","url":null,"abstract":"Enterprise Architecture Management (EAM) is widely used in the public sector and is increasingly understood as a driver of digital transformation. After reviewing the current literature on the EAM Services in the public sector, this article reports on experiences in the realignment of the EAM services in Deutsche Rentenversicherung (DRV - German pension insurance), presents an approach supported by Value Proposition Canvas (VPC), and details the EAM services that were designed using the method.","PeriodicalId":93549,"journal":{"name":"EPiC series in computing","volume":"1 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2022-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"69423129","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：