Pub Date : 2020-04-01DOI: 10.1109/SIEDS49339.2020.9106639
Vidhi Gupta, Guangda Zhu, Andi Yu, Donald E. Brown
Contact centers provide customer interaction support to numerous organizations. In 2017, the contact center industry generated $200 billion in revenue worldwide, contributing to a significant proportion of market share, and yet businesses lost $75 billion due to poor customer satisfaction. Around 48% of consumers prefer using phones as their mode of communication with contact centers. Analysis of these calls can give insights into customer views and help businesses improve their customer engagement. To understand the structure and flow of the conversation, the conversation transcript can be segmented into meaningful sections such as “greeting exchange” “problem description” and “problem resolution”, to name a few. In this paper, we present a comparative study of various unsupervised methods of dialogue segmentation. We choose three classic unsupervised text segmentation techniques: TextTiling, TopicTiling, and Content Vector Segmentation, and evaluate their performance on 50 manually labeled dialogue conversation transcripts. The transcripts used span across contact center calls, live chat, interactions with chat-bots and talk show conversations. Additionally, we build on the TextTiling algorithm by incorporating semantic word embeddings for text representation. We show that this modification outperforms the three benchmarked approaches with a mean Pk value of 0.31, indicating that 69% of the boundaries are identified accurately at an average.
{"title":"A Comparative Study of the Performance of Unsupervised Text Segmentation Techniques on Dialogue Transcripts","authors":"Vidhi Gupta, Guangda Zhu, Andi Yu, Donald E. Brown","doi":"10.1109/SIEDS49339.2020.9106639","DOIUrl":"https://doi.org/10.1109/SIEDS49339.2020.9106639","url":null,"abstract":"Contact centers provide customer interaction support to numerous organizations. In 2017, the contact center industry generated $200 billion in revenue worldwide, contributing to a significant proportion of market share, and yet businesses lost $75 billion due to poor customer satisfaction. Around 48% of consumers prefer using phones as their mode of communication with contact centers. Analysis of these calls can give insights into customer views and help businesses improve their customer engagement. To understand the structure and flow of the conversation, the conversation transcript can be segmented into meaningful sections such as “greeting exchange” “problem description” and “problem resolution”, to name a few. In this paper, we present a comparative study of various unsupervised methods of dialogue segmentation. We choose three classic unsupervised text segmentation techniques: TextTiling, TopicTiling, and Content Vector Segmentation, and evaluate their performance on 50 manually labeled dialogue conversation transcripts. The transcripts used span across contact center calls, live chat, interactions with chat-bots and talk show conversations. Additionally, we build on the TextTiling algorithm by incorporating semantic word embeddings for text representation. We show that this modification outperforms the three benchmarked approaches with a mean Pk value of 0.31, indicating that 69% of the boundaries are identified accurately at an average.","PeriodicalId":331495,"journal":{"name":"2020 Systems and Information Engineering Design Symposium (SIEDS)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128713585","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 : 2020-04-01DOI: 10.1109/SIEDS49339.2020.9106690
Aamil Rastogi, Smriti Sridhar, R. Gupta
The increase in the crime rate numbers and a rise in the need to find better solutions to handle information about criminality is affected by the ever-changing socio-economical order of the world. Despite the number of solutions implemented for reducing crime (against women), cities continue to have an unsafe environment. The prime drawback lies in the inability to provide a prompt response in real-time when in danger. Thus, the effective utilization of technology in public safety management is important. The present state of the art solutions focus on technological innovations with limited human intervention and are insufficient in ensuring the safety of the women as and when required. To dig deeper into the root cause of preventing a crime from occurring in a particular place, it is vital to analyze the parameters and factors contributing to the crime in a community. This research applies the Information Communication Technologies (ICT) along with harnessing big data tools to identify crime hotspots and patterns. After a comprehensive literature review, it has been noted that there are different social-economic factors affecting the crime in an area. The proposed work aims to integrate the socio-economic attributes leading to increasing crime against women. Interpolation strategies used for thematic maps generation also play a major role in predicting and studying the area affected by a crime. This research initially identifies the various social-economical parameters that affect crime against women. Some of them to mention include unemployment, illiteracy, population, sex ratio, traffic, age, no. of schools, and location of liquor shops. Subsequently, a comparison of major interpolation methods used in crime mapping: Inverse Distance Weighted (IDW), Kriging, and Spline are formulated to understand the overall contribution of socio-economic factors on the crime thematic map to further ascertain if one parameter poses substantially more important than the other. The comparison of different Interpolation techniques used in pixel by pixel error analysis on high definition satellite images of the crime site, of resolution as high as 2.5m x 2.5m, is created using visualization libraries like Matplotlib and Seaborn. Finally, the thematic maps are created using the best Interpolation technique chosen and help in predicting the pattern of the crime. The proposed framework developed using Geographic Information System (GIS) based visualization and big data tools for crime mapping can then be applied in the development of user interactive platforms and designing safety strategies to help the needy in real-time. To validate the methodology, a case study is performed with real data, in the Jhunjhunu district of Rajasthan, India.
世界上不断变化的社会经济秩序影响到犯罪率数字的增加和需要找到更好的解决办法来处理有关犯罪的信息。尽管实施了许多减少(针对妇女的)犯罪的解决办法,但城市的环境仍然不安全。主要的缺点是在遇到危险时无法及时作出反应。因此,技术在公共安全管理中的有效利用具有重要意义。目前最先进的解决办法集中于技术革新,人为干预有限,在必要时确保妇女的安全方面是不够的。为了更深入地挖掘防止犯罪在特定地方发生的根本原因,分析导致社区犯罪的参数和因素至关重要。本研究应用信息通信技术(ICT)以及利用大数据工具来识别犯罪热点和模式。在全面的文献综述之后,我们注意到一个地区的犯罪受到不同的社会经济因素的影响。拟议的工作旨在综合导致针对妇女的犯罪增加的社会经济因素。用于专题地图生成的插值策略在预测和研究受犯罪影响的区域方面也起着重要作用。这项研究最初确定了影响针对妇女犯罪的各种社会经济参数。其中包括失业,文盲,人口,性别比例,交通,年龄,没有。学校的位置,酒类商店的位置。随后,对犯罪地图中使用的主要插值方法进行了比较:逆距离加权法(IDW)、克里格法(Kriging)和样条法(Spline),以了解社会经济因素对犯罪主题地图的总体贡献,从而进一步确定一个参数是否比另一个参数更重要。利用Matplotlib和Seaborn等可视化库,对犯罪现场分辨率高达2.5m x 2.5m的高清卫星图像进行逐像素误差分析时使用的不同插值技术进行比较。最后,使用选择的最佳插值技术创建主题地图,并帮助预测犯罪模式。利用基于地理信息系统(GIS)的可视化和大数据工具开发的犯罪地图框架,可以应用于开发用户交互平台和设计安全策略,实时帮助有需要的人。为了验证该方法,在印度拉贾斯坦邦Jhunjhunu地区进行了实际数据的案例研究。
{"title":"Comparison of Different Spatial Interpolation Techniques to Thematic Mapping of Socio-Economic Causes of Crime Against Women","authors":"Aamil Rastogi, Smriti Sridhar, R. Gupta","doi":"10.1109/SIEDS49339.2020.9106690","DOIUrl":"https://doi.org/10.1109/SIEDS49339.2020.9106690","url":null,"abstract":"The increase in the crime rate numbers and a rise in the need to find better solutions to handle information about criminality is affected by the ever-changing socio-economical order of the world. Despite the number of solutions implemented for reducing crime (against women), cities continue to have an unsafe environment. The prime drawback lies in the inability to provide a prompt response in real-time when in danger. Thus, the effective utilization of technology in public safety management is important. The present state of the art solutions focus on technological innovations with limited human intervention and are insufficient in ensuring the safety of the women as and when required. To dig deeper into the root cause of preventing a crime from occurring in a particular place, it is vital to analyze the parameters and factors contributing to the crime in a community. This research applies the Information Communication Technologies (ICT) along with harnessing big data tools to identify crime hotspots and patterns. After a comprehensive literature review, it has been noted that there are different social-economic factors affecting the crime in an area. The proposed work aims to integrate the socio-economic attributes leading to increasing crime against women. Interpolation strategies used for thematic maps generation also play a major role in predicting and studying the area affected by a crime. This research initially identifies the various social-economical parameters that affect crime against women. Some of them to mention include unemployment, illiteracy, population, sex ratio, traffic, age, no. of schools, and location of liquor shops. Subsequently, a comparison of major interpolation methods used in crime mapping: Inverse Distance Weighted (IDW), Kriging, and Spline are formulated to understand the overall contribution of socio-economic factors on the crime thematic map to further ascertain if one parameter poses substantially more important than the other. The comparison of different Interpolation techniques used in pixel by pixel error analysis on high definition satellite images of the crime site, of resolution as high as 2.5m x 2.5m, is created using visualization libraries like Matplotlib and Seaborn. Finally, the thematic maps are created using the best Interpolation technique chosen and help in predicting the pattern of the crime. The proposed framework developed using Geographic Information System (GIS) based visualization and big data tools for crime mapping can then be applied in the development of user interactive platforms and designing safety strategies to help the needy in real-time. To validate the methodology, a case study is performed with real data, in the Jhunjhunu district of Rajasthan, India.","PeriodicalId":331495,"journal":{"name":"2020 Systems and Information Engineering Design Symposium (SIEDS)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115839907","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 : 2020-04-01DOI: 10.1109/SIEDS49339.2020.9106634
Kevin Malloy, S. Kausch, Aneesh Sandhir
Over one million Americans are currently living with HIV. The Ryan White HIV/AIDS Program (RWHAP) provides funding for HIV medical care and medications for people living with HIV. The RWHAP program received 2.34 billion dollars in 2018 and the Ending the HIV Epidemic initiative was awarded 117 million dollars in 2020. However, even with increased funding, geographic barriers to accessing HIV care can prevent people from obtaining treatment. Additionally, the impact of insurance status (none, Medicaid, Affordable Care Act plans) on drive times to HIV care is not well understood. Geographic access to RWHAP clinics in the contiguous United States was examined. Using spatial analysis techniques, the duration of drive time from the center of every county equivalent to the nearest accessible RWHAP clinic was measured. Counties were characterized in terms of social determinants of health and HIVrelated variables and their associations with access to HIV care were examined.The effect of insurance status on drive times was analyzed in order to measure its impact by being uninsured, enrolled in Medicaid, or enrolled in either the least or most expensive Affordable Care Act plans.Four hundred twenty-seven RWHAP locations were identified with a median county-level drive time of 64.6 minutes (interquartile range (IQR) 40.9-97.9) for counties with five or more diagnosed HIV cases. The median drive time for Medicaid access was 69.3 minutes (IQR 42.2-106.0), with some states impacted more than others. These findings were used to make specific policy recommendations to improve access and reduce barriers to HIV care.
{"title":"Geographic Access to HIV Care","authors":"Kevin Malloy, S. Kausch, Aneesh Sandhir","doi":"10.1109/SIEDS49339.2020.9106634","DOIUrl":"https://doi.org/10.1109/SIEDS49339.2020.9106634","url":null,"abstract":"Over one million Americans are currently living with HIV. The Ryan White HIV/AIDS Program (RWHAP) provides funding for HIV medical care and medications for people living with HIV. The RWHAP program received 2.34 billion dollars in 2018 and the Ending the HIV Epidemic initiative was awarded 117 million dollars in 2020. However, even with increased funding, geographic barriers to accessing HIV care can prevent people from obtaining treatment. Additionally, the impact of insurance status (none, Medicaid, Affordable Care Act plans) on drive times to HIV care is not well understood. Geographic access to RWHAP clinics in the contiguous United States was examined. Using spatial analysis techniques, the duration of drive time from the center of every county equivalent to the nearest accessible RWHAP clinic was measured. Counties were characterized in terms of social determinants of health and HIVrelated variables and their associations with access to HIV care were examined.The effect of insurance status on drive times was analyzed in order to measure its impact by being uninsured, enrolled in Medicaid, or enrolled in either the least or most expensive Affordable Care Act plans.Four hundred twenty-seven RWHAP locations were identified with a median county-level drive time of 64.6 minutes (interquartile range (IQR) 40.9-97.9) for counties with five or more diagnosed HIV cases. The median drive time for Medicaid access was 69.3 minutes (IQR 42.2-106.0), with some states impacted more than others. These findings were used to make specific policy recommendations to improve access and reduce barriers to HIV care.","PeriodicalId":331495,"journal":{"name":"2020 Systems and Information Engineering Design Symposium (SIEDS)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115598217","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 : 2020-04-01DOI: 10.1109/sieds49339.2020.9106629
Siddharth Suresh, Devan Visvalingam, Adonis Lu, Briana K. Wright
Analyzing customer attrition in the retail banking industry allows banks to quantify the likelihood of a customer closing their account. With the onset of online banking services, it has become important to both understand the latent behavioral patterns behind attrition and predict the event of attrition well before losing a customer. Presently, attrition models measure hard attrition, the event of a customer closing their account. By introducing a new latent probabilistic response variable, soft attrition, we aim to identify customers that tend towards attrition, which (i) increases the comprehensiveness of the customer base that is likely to churn, (ii) improves capability of predicting attrition events early, and (iii) helps identify key features associated with attrition. This paper introduces and evaluates methods that help redefine the attrition response variable and proposes techniques that improve on the existing attrition models, specifically in the retail banking industry.
{"title":"Evaluating and Improving Attrition Models for the Retail Banking Industry","authors":"Siddharth Suresh, Devan Visvalingam, Adonis Lu, Briana K. Wright","doi":"10.1109/sieds49339.2020.9106629","DOIUrl":"https://doi.org/10.1109/sieds49339.2020.9106629","url":null,"abstract":"Analyzing customer attrition in the retail banking industry allows banks to quantify the likelihood of a customer closing their account. With the onset of online banking services, it has become important to both understand the latent behavioral patterns behind attrition and predict the event of attrition well before losing a customer. Presently, attrition models measure hard attrition, the event of a customer closing their account. By introducing a new latent probabilistic response variable, soft attrition, we aim to identify customers that tend towards attrition, which (i) increases the comprehensiveness of the customer base that is likely to churn, (ii) improves capability of predicting attrition events early, and (iii) helps identify key features associated with attrition. This paper introduces and evaluates methods that help redefine the attrition response variable and proposes techniques that improve on the existing attrition models, specifically in the retail banking industry.","PeriodicalId":331495,"journal":{"name":"2020 Systems and Information Engineering Design Symposium (SIEDS)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114218925","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 : 2020-04-01DOI: 10.1109/SIEDS49339.2020.9106631
John H. Mott, Bhavana Kotla
The recent increase in school bus passing accidents leading to injuries to and deaths of children has sparked nationwide concern; hence, the development of a comprehensive solution to minimize the probabilities of such accidents is essential. Ignoring active stop signs and crossing arms on school buses that are boarding and deboarding children at stops is a serious traffic violation, but it is difficult to prosecute drivers who are guilty of these offenses due to the lack of surveillance and monitoring systems that can provide critical violation information to law enforcement agencies. This study aims to address the issue of illegal passing of school buses prevailing in small towns and cities where lack of sufficient oversight exists. A detection system incorporating a solid-state LiDAR unit and a dashcam, both of which are controlled by a Raspberry Pi computer, was designed. The primary function of the system is to capture an image of the license plate of the violating vehicle and make that data available to law enforcement agencies, enabling those agencies to take appropriate enforcement action, which in turn will serve as a deterrent to mitigate future accidents. Several state legislative bodies have passed related bills and have urged researchers to find solutions to address the issue. This detection system achieves two keys goals: reducing overall cost of system implementation and reducing video review time. It is a potential component for a comprehensive solution to the school bus passing problem.
{"title":"Design and Validation of a School Bus Passing Detection System Based on Solid-State LiDAR","authors":"John H. Mott, Bhavana Kotla","doi":"10.1109/SIEDS49339.2020.9106631","DOIUrl":"https://doi.org/10.1109/SIEDS49339.2020.9106631","url":null,"abstract":"The recent increase in school bus passing accidents leading to injuries to and deaths of children has sparked nationwide concern; hence, the development of a comprehensive solution to minimize the probabilities of such accidents is essential. Ignoring active stop signs and crossing arms on school buses that are boarding and deboarding children at stops is a serious traffic violation, but it is difficult to prosecute drivers who are guilty of these offenses due to the lack of surveillance and monitoring systems that can provide critical violation information to law enforcement agencies. This study aims to address the issue of illegal passing of school buses prevailing in small towns and cities where lack of sufficient oversight exists. A detection system incorporating a solid-state LiDAR unit and a dashcam, both of which are controlled by a Raspberry Pi computer, was designed. The primary function of the system is to capture an image of the license plate of the violating vehicle and make that data available to law enforcement agencies, enabling those agencies to take appropriate enforcement action, which in turn will serve as a deterrent to mitigate future accidents. Several state legislative bodies have passed related bills and have urged researchers to find solutions to address the issue. This detection system achieves two keys goals: reducing overall cost of system implementation and reducing video review time. It is a potential component for a comprehensive solution to the school bus passing problem.","PeriodicalId":331495,"journal":{"name":"2020 Systems and Information Engineering Design Symposium (SIEDS)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128145047","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 : 2020-04-01DOI: 10.1109/SIEDS49339.2020.9106643
Alexandria Bellas, Stefawn Perrin, Brandon Malone, Kaytlin Rogers, Gale M. Lucas, Elizabeth Phillips, Chad C. Tossell, E. D. Visser
Conflicts may arise at any time during military debriefing meetings, especially in high intensity deployed settings. When such conflicts arise, it takes time to get everyone back into a receptive state of mind so that they engage in reflective discussion rather than unproductive arguing. It has been proposed by some that the use of social robots equipped with social abilities such as emotion regulation through rapport building may help to deescalate these situations to facilitate critical operational decisions. However, in military settings, the same AI agent used in the pre-brief of a mission may not be the same one used in the debrief. The purpose of this study was to determine whether a brief rapport-building session with a social robot could create a connection between a human and a robot agent, and whether consistency in the embodiment of the robot agent was necessary for maintaining this connection once formed. We report the results of a pilot study conducted at the United States Air Force Academy which simulated a military mission (i.e., Gravity and Strike). Participants’ connection with the agent, sense of trust, and overall likeability revealed that early rapport building can be beneficial for military missions.
{"title":"Rapport Building with Social Robots as a Method for Improving Mission Debriefing in Human-Robot Teams","authors":"Alexandria Bellas, Stefawn Perrin, Brandon Malone, Kaytlin Rogers, Gale M. Lucas, Elizabeth Phillips, Chad C. Tossell, E. D. Visser","doi":"10.1109/SIEDS49339.2020.9106643","DOIUrl":"https://doi.org/10.1109/SIEDS49339.2020.9106643","url":null,"abstract":"Conflicts may arise at any time during military debriefing meetings, especially in high intensity deployed settings. When such conflicts arise, it takes time to get everyone back into a receptive state of mind so that they engage in reflective discussion rather than unproductive arguing. It has been proposed by some that the use of social robots equipped with social abilities such as emotion regulation through rapport building may help to deescalate these situations to facilitate critical operational decisions. However, in military settings, the same AI agent used in the pre-brief of a mission may not be the same one used in the debrief. The purpose of this study was to determine whether a brief rapport-building session with a social robot could create a connection between a human and a robot agent, and whether consistency in the embodiment of the robot agent was necessary for maintaining this connection once formed. We report the results of a pilot study conducted at the United States Air Force Academy which simulated a military mission (i.e., Gravity and Strike). Participants’ connection with the agent, sense of trust, and overall likeability revealed that early rapport building can be beneficial for military missions.","PeriodicalId":331495,"journal":{"name":"2020 Systems and Information Engineering Design Symposium (SIEDS)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126529282","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 : 2020-04-01DOI: 10.1109/SIEDS49339.2020.9106657
Chuyang Yang, Zachary A. Marshall, John H. Mott
Flight delays in the U. S. National Airspace System (NAS) present a fundamental challenge to capacity growth under ever-increasing traffic volumes, and lead to significant financial burdens that reverberate across a multitude of aviation industry stakeholders. Roughly 20% of passengers’ total travel time is due to such delays, causing $35 billion annually in lost revenue and impacting not only the airline industry, but the retail, lodging, restaurant, and tourism industries, as well. The Federal Aviation Administration’s effort in aiding decision-making at airports is readily apparent in the Next Generation Air Traffic Control (NextGen) System’s System-Wide Information Management (SWIM) program, and in-flight delay information from the FAA Air Traffic Control System Command Center (ATCSCC). Academic researchers are concurrently developing various algorithms to predict flight delays that include advanced statistics, machine learning, and graph theory using various network topologies. Other stakeholders have initiated delay prediction methods to adjust their operational schedules. This suggests an opportunity to centralize, validate, and integrate the various delay prediction methods under development; furthermore, these methods are limited in scope with regard to geography, operators, and efficacy.The authors propose here a platform supporting the FAA’s Collaborative Decision-Making (CDM) process with the intent of reducing flight delays in the NAS. Building upon existing deep learning algorithms and utilizing the NextGen SWIM program, this research suggests a central delay prediction platform suited to the complex and dynamic needs of America’s airport infrastructure. assessments of risks and sustainability of the proposed platform are presented. The authors interviewed experts in industry and academic fields related to aviation and information technology, and used the information obtained to refine the model. It is anticipated that this model will accurately produce location-specific departure and arrival delay forecasts that can further be integrated into the CDM and Ground Delay Program (GDP) initiatives.
{"title":"A Novel Integration Platform to Reduce Flight Delays in the National Airspace System","authors":"Chuyang Yang, Zachary A. Marshall, John H. Mott","doi":"10.1109/SIEDS49339.2020.9106657","DOIUrl":"https://doi.org/10.1109/SIEDS49339.2020.9106657","url":null,"abstract":"Flight delays in the U. S. National Airspace System (NAS) present a fundamental challenge to capacity growth under ever-increasing traffic volumes, and lead to significant financial burdens that reverberate across a multitude of aviation industry stakeholders. Roughly 20% of passengers’ total travel time is due to such delays, causing $35 billion annually in lost revenue and impacting not only the airline industry, but the retail, lodging, restaurant, and tourism industries, as well. The Federal Aviation Administration’s effort in aiding decision-making at airports is readily apparent in the Next Generation Air Traffic Control (NextGen) System’s System-Wide Information Management (SWIM) program, and in-flight delay information from the FAA Air Traffic Control System Command Center (ATCSCC). Academic researchers are concurrently developing various algorithms to predict flight delays that include advanced statistics, machine learning, and graph theory using various network topologies. Other stakeholders have initiated delay prediction methods to adjust their operational schedules. This suggests an opportunity to centralize, validate, and integrate the various delay prediction methods under development; furthermore, these methods are limited in scope with regard to geography, operators, and efficacy.The authors propose here a platform supporting the FAA’s Collaborative Decision-Making (CDM) process with the intent of reducing flight delays in the NAS. Building upon existing deep learning algorithms and utilizing the NextGen SWIM program, this research suggests a central delay prediction platform suited to the complex and dynamic needs of America’s airport infrastructure. assessments of risks and sustainability of the proposed platform are presented. The authors interviewed experts in industry and academic fields related to aviation and information technology, and used the information obtained to refine the model. It is anticipated that this model will accurately produce location-specific departure and arrival delay forecasts that can further be integrated into the CDM and Ground Delay Program (GDP) initiatives.","PeriodicalId":331495,"journal":{"name":"2020 Systems and Information Engineering Design Symposium (SIEDS)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121700235","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 : 2020-04-01DOI: 10.1109/SIEDS49339.2020.9106685
Avery Tang, Timothy (Jun) Lu, Z. Lynch, Oliver Schaer, Stephen Adams
When it comes to making promotions, companies rely upon a variety of metrics and rating systems to support their decisions. However, are they looking at the most important metrics and more broadly, how should they identify employees to promote? The literature predominantly focuses on the measurement of performance, but businesses also need instruments that can predict management potential for promotional decision-making. This paper utilizes the data contained in the Human Resources Information System (HRIS) of a company to analyze drivers of potential for promotion among a sample of its workers. Numerous prior studies have been conducted of human resource variables in a variety of organizations. These studies share in common the use of linear models to report which explanatory variables are statistically significant determinants of the dependent variable – in most cases the performance of employees with a focus on the individual’s output. What they do not deliver, and what this study provides, in addition to regression studies on employee performance, is an analysis of the drivers of promotion potential for management roles. The perspective of our analysis diverges from others in that its primary focus is to identify future leaders of a company rather than identifying strong individual contributors. The methods used consist of basic statistical procedures, multiple classification methods and graph theory analysis. In our study of managerial potential drivers, the logistic regression model performs with the best predictive accuracy and recognizes which factors in a manager reveals leadership potential. In our study of promotion potential from a teamwork perspective, we show that graph network-based methods adapt well to employee data containing several bilateral relationships while preserving the hierarchy of an organization and providing defensible accuracy.
{"title":"Enhancing Promotion Decisions using Classification and Network-based Methods","authors":"Avery Tang, Timothy (Jun) Lu, Z. Lynch, Oliver Schaer, Stephen Adams","doi":"10.1109/SIEDS49339.2020.9106685","DOIUrl":"https://doi.org/10.1109/SIEDS49339.2020.9106685","url":null,"abstract":"When it comes to making promotions, companies rely upon a variety of metrics and rating systems to support their decisions. However, are they looking at the most important metrics and more broadly, how should they identify employees to promote? The literature predominantly focuses on the measurement of performance, but businesses also need instruments that can predict management potential for promotional decision-making. This paper utilizes the data contained in the Human Resources Information System (HRIS) of a company to analyze drivers of potential for promotion among a sample of its workers. Numerous prior studies have been conducted of human resource variables in a variety of organizations. These studies share in common the use of linear models to report which explanatory variables are statistically significant determinants of the dependent variable – in most cases the performance of employees with a focus on the individual’s output. What they do not deliver, and what this study provides, in addition to regression studies on employee performance, is an analysis of the drivers of promotion potential for management roles. The perspective of our analysis diverges from others in that its primary focus is to identify future leaders of a company rather than identifying strong individual contributors. The methods used consist of basic statistical procedures, multiple classification methods and graph theory analysis. In our study of managerial potential drivers, the logistic regression model performs with the best predictive accuracy and recognizes which factors in a manager reveals leadership potential. In our study of promotion potential from a teamwork perspective, we show that graph network-based methods adapt well to employee data containing several bilateral relationships while preserving the hierarchy of an organization and providing defensible accuracy.","PeriodicalId":331495,"journal":{"name":"2020 Systems and Information Engineering Design Symposium (SIEDS)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121125639","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 : 2020-04-01DOI: 10.1109/SIEDS49339.2020.9106648
Erin K. Barrett, Cameron M. Fard, Hannah N. Katinas, Charles V. Moens, Lauren E. Perry, Blake E. Ruddy, Shalin S Shah, Ian Tucker, Tucker J. Wilson, Mark Rucker, Lihua Cai, Laura E. Barnes, M. Boukhechba
Smartphones can collect millions of data points from each of its users daily, contributing to a significant change in how the healthcare community approaches health monitoring. This paper provides a framework for how smartphone sensor data can be collected, cleaned, stored, and modeled to effectively predict human states as a step towards health monitoring. To develop robust contextual models, a three-week study was conducted to collect data through a mobile crowdsensing application named Sensus. In this study, participants used multiple sensing strategies, ranging from infrequent sampling to continuous sampling, to determine the effect each has on data integrity and battery life. For a future study, a dynamic data collection strategy was developed that uses a machine learning model trained on existing data collected from 220 participants to forecast when a smartphone will be active and trigger sensor sampling accordingly. Results of this study include 1) extraction of model features that deliver maximized data quality with minimized battery consumption as compared to pre-existing baseline models, 2) implementation of context-driven modeling of user smartphone data on user's contextual environment, and 3) customization of a time-series database for optimized data queries used in metadata visualizations. The adaptive sensing models produced could be used in future large population studies that efficiently examine patterns of behavior in multiple individuals over extended periods to identify disease indicators present in an average user’s daily life.
{"title":"Mobile Sensing: Leveraging Machine Learning for Efficient Human Behavior Modeling","authors":"Erin K. Barrett, Cameron M. Fard, Hannah N. Katinas, Charles V. Moens, Lauren E. Perry, Blake E. Ruddy, Shalin S Shah, Ian Tucker, Tucker J. Wilson, Mark Rucker, Lihua Cai, Laura E. Barnes, M. Boukhechba","doi":"10.1109/SIEDS49339.2020.9106648","DOIUrl":"https://doi.org/10.1109/SIEDS49339.2020.9106648","url":null,"abstract":"Smartphones can collect millions of data points from each of its users daily, contributing to a significant change in how the healthcare community approaches health monitoring. This paper provides a framework for how smartphone sensor data can be collected, cleaned, stored, and modeled to effectively predict human states as a step towards health monitoring. To develop robust contextual models, a three-week study was conducted to collect data through a mobile crowdsensing application named Sensus. In this study, participants used multiple sensing strategies, ranging from infrequent sampling to continuous sampling, to determine the effect each has on data integrity and battery life. For a future study, a dynamic data collection strategy was developed that uses a machine learning model trained on existing data collected from 220 participants to forecast when a smartphone will be active and trigger sensor sampling accordingly. Results of this study include 1) extraction of model features that deliver maximized data quality with minimized battery consumption as compared to pre-existing baseline models, 2) implementation of context-driven modeling of user smartphone data on user's contextual environment, and 3) customization of a time-series database for optimized data queries used in metadata visualizations. The adaptive sensing models produced could be used in future large population studies that efficiently examine patterns of behavior in multiple individuals over extended periods to identify disease indicators present in an average user’s daily life.","PeriodicalId":331495,"journal":{"name":"2020 Systems and Information Engineering Design Symposium (SIEDS)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129511511","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 : 2020-04-01DOI: 10.1109/SIEDS49339.2020.9106692
M. Kuchta, Christopher Pufko, Charles Rowe, Scott Stoessel, J. Walsh, V. Lakshmi
The Mekong River region’s long-term social and economic sustainability is being threatened by the growing development of hydropower and its impacts on the river, surrounding populations, and vital industries. In this study we have analyzed these unintended impacts through data analysis in hopes of quantifying trends associated with the rapid hydropower development. It is important to consider the human and social dimensions of hydropower in the area as the dams’ effects trickle down to the natives of the Mekong Region, the river itself, and all other life dependent on it. We conducted our research by utilizing data sets and surveys released by certain organizations such as the FAO (Food and Agriculture Organization), the WB (World Bank), and CGIAR (Consultative Group for Agricultural Research) International to develop a basis for drawing conclusions.In this study, we segment the analysis into five sectors: hydropower, agriculture, fisheries and aquaculture, economy, and land use. We then correlate dam implementation and hydroelectric capacity with impacts to the Mekong River Basin. Through our research, we expect to find quantifiable correlations between the increased development of hydropower and the resulting impacts on the Mekong’s inhabitants and the region’s overall well-being.
{"title":"Understanding the Land Use and Water Systems of the Mekong River","authors":"M. Kuchta, Christopher Pufko, Charles Rowe, Scott Stoessel, J. Walsh, V. Lakshmi","doi":"10.1109/SIEDS49339.2020.9106692","DOIUrl":"https://doi.org/10.1109/SIEDS49339.2020.9106692","url":null,"abstract":"The Mekong River region’s long-term social and economic sustainability is being threatened by the growing development of hydropower and its impacts on the river, surrounding populations, and vital industries. In this study we have analyzed these unintended impacts through data analysis in hopes of quantifying trends associated with the rapid hydropower development. It is important to consider the human and social dimensions of hydropower in the area as the dams’ effects trickle down to the natives of the Mekong Region, the river itself, and all other life dependent on it. We conducted our research by utilizing data sets and surveys released by certain organizations such as the FAO (Food and Agriculture Organization), the WB (World Bank), and CGIAR (Consultative Group for Agricultural Research) International to develop a basis for drawing conclusions.In this study, we segment the analysis into five sectors: hydropower, agriculture, fisheries and aquaculture, economy, and land use. We then correlate dam implementation and hydroelectric capacity with impacts to the Mekong River Basin. Through our research, we expect to find quantifiable correlations between the increased development of hydropower and the resulting impacts on the Mekong’s inhabitants and the region’s overall well-being.","PeriodicalId":331495,"journal":{"name":"2020 Systems and Information Engineering Design Symposium (SIEDS)","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2020-04-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131057332","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}