The stability of geomagnetism can be used as an indoor positioning fingerprint mark, which has good precision and applicability, therefore, the study of geomagnetism has become an emerging direction of indoor positioning in recent years. In the existing research, the use of geomagnetism mostly follows the idea of building a geomagnetic fingerprint map and then real-time similarity matching online. However, there are still many new ideas for the application of geomagnetism. In this paper, based on the research of the existing geomagnetic framework, two optimizations are made. One is a geomagnetic data migration(GDM) model based on data similarity. The model is mainly for the difference of the built-in geomagnetic sensor of different mobile phone models. When the indoor environment does not change greatly, the standard geomagnetic acquisition sensor is used for one acquisition, other types of mobile phones use the similarity matching model to calculate the geomagnetic fingerprint map without first acquiring geomagnetism in advance, thereby performing indoor positioning. The other is a step counting optimization model based on geomagnetic assisted accelerometer(GASC), a geomagnetic dynamic threshold method is proposed by data mining of shaking mobile phone and geomagnetic variation trend while walking. By combining with the traditional accelerometer threshold method model, the pseudo step counting recognition ability is improved. The experimental results show that the optimized model performs better anti-interference in the case of shaking the mobile phone.
{"title":"The Improvement of Traditional Indoor Localization Model Using Magnetic Field Based on Smartphone","authors":"Shanzhi Gu, Ruyi Yao, L. Lan, Chao Guo, Feng Gao, Chuanfu Xu","doi":"10.1109/ISKE47853.2019.9170444","DOIUrl":"https://doi.org/10.1109/ISKE47853.2019.9170444","url":null,"abstract":"The stability of geomagnetism can be used as an indoor positioning fingerprint mark, which has good precision and applicability, therefore, the study of geomagnetism has become an emerging direction of indoor positioning in recent years. In the existing research, the use of geomagnetism mostly follows the idea of building a geomagnetic fingerprint map and then real-time similarity matching online. However, there are still many new ideas for the application of geomagnetism. In this paper, based on the research of the existing geomagnetic framework, two optimizations are made. One is a geomagnetic data migration(GDM) model based on data similarity. The model is mainly for the difference of the built-in geomagnetic sensor of different mobile phone models. When the indoor environment does not change greatly, the standard geomagnetic acquisition sensor is used for one acquisition, other types of mobile phones use the similarity matching model to calculate the geomagnetic fingerprint map without first acquiring geomagnetism in advance, thereby performing indoor positioning. The other is a step counting optimization model based on geomagnetic assisted accelerometer(GASC), a geomagnetic dynamic threshold method is proposed by data mining of shaking mobile phone and geomagnetic variation trend while walking. By combining with the traditional accelerometer threshold method model, the pseudo step counting recognition ability is improved. The experimental results show that the optimized model performs better anti-interference in the case of shaking the mobile phone.","PeriodicalId":399084,"journal":{"name":"2019 IEEE 14th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127654429","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 : 2019-11-01DOI: 10.1109/iske47853.2019.9170281
Mi Swe Zar Thu, Ei Chaw Htoon
Smart mobile devices are essential for technology trend of modern lifestyles. Mobile applications are getting more diverse and complex with an increasing use of mobile devices. At mobile environment, resource limitation is irresistible extremely; it is a key challenge and can impact on mobile computing performance. Mobile Cloud computing (MCC) is one of the eventual gold computing executions of rich mobile application on an abundance of mobile devices. Limited computational power, storage, and energy are necessary for hardware limitations of offloading. So, the proposed system aims to avoid limitation of devices starting in case of intensive tasks by using dynamic computation offloading of Markov process. The system reduces energy consumption of resource hungry device as the way of taking decision to offload the computing intensive tasks to the remote cloud by the result of our cost model. The experiment shows that the proposed offloading decision solver can reduce not only computing time but also battery usage of mobile device for face detection application compared to the dynamic off loading MAUI decision framework.
{"title":"Markov Based Computational Tasks Offloading Decision for Face Detection","authors":"Mi Swe Zar Thu, Ei Chaw Htoon","doi":"10.1109/iske47853.2019.9170281","DOIUrl":"https://doi.org/10.1109/iske47853.2019.9170281","url":null,"abstract":"Smart mobile devices are essential for technology trend of modern lifestyles. Mobile applications are getting more diverse and complex with an increasing use of mobile devices. At mobile environment, resource limitation is irresistible extremely; it is a key challenge and can impact on mobile computing performance. Mobile Cloud computing (MCC) is one of the eventual gold computing executions of rich mobile application on an abundance of mobile devices. Limited computational power, storage, and energy are necessary for hardware limitations of offloading. So, the proposed system aims to avoid limitation of devices starting in case of intensive tasks by using dynamic computation offloading of Markov process. The system reduces energy consumption of resource hungry device as the way of taking decision to offload the computing intensive tasks to the remote cloud by the result of our cost model. The experiment shows that the proposed offloading decision solver can reduce not only computing time but also battery usage of mobile device for face detection application compared to the dynamic off loading MAUI decision framework.","PeriodicalId":399084,"journal":{"name":"2019 IEEE 14th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114725799","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 : 2019-11-01DOI: 10.1109/ISKE47853.2019.9170443
Wenhui Yu, Zhengrong Wu, Xinye Bao, L. Yin, Yaowen Liang, Yan He
Geographic Information System (GIS) is the basic data management and visualization platform for transmission and distribution power network planning, operation scheduling and repair decision support systems. GIS systems based on SQL database can’t meet the real-time requirements of massive data processing and large-scale concurrent topology analysis of distribution networks. An object-oriented in-memory database is introduced, which uses partitioning and paging storage technology for efficient caching and retrieval of large-scale grid topology models. Parallel processing techniques based on data partitions and task scheduling queues are developed, which enables the parallel executions of multi-user requests of topology tracing(reading) and switch open-close operations(writing). Further, for the conflicting write requests across partitions, a parent-child task queue is introduced. In the stress test of the on-line system of a provincial company with more than 10 million power grid equipment, response time less than 0.2 seconds is observed, under the load of more than 400 topology analysis requests per second per server.
{"title":"Application of In-Memory Database in Concurrent Topology Analysis of GIS Systems for Large-Scale Distribution Power Grids","authors":"Wenhui Yu, Zhengrong Wu, Xinye Bao, L. Yin, Yaowen Liang, Yan He","doi":"10.1109/ISKE47853.2019.9170443","DOIUrl":"https://doi.org/10.1109/ISKE47853.2019.9170443","url":null,"abstract":"Geographic Information System (GIS) is the basic data management and visualization platform for transmission and distribution power network planning, operation scheduling and repair decision support systems. GIS systems based on SQL database can’t meet the real-time requirements of massive data processing and large-scale concurrent topology analysis of distribution networks. An object-oriented in-memory database is introduced, which uses partitioning and paging storage technology for efficient caching and retrieval of large-scale grid topology models. Parallel processing techniques based on data partitions and task scheduling queues are developed, which enables the parallel executions of multi-user requests of topology tracing(reading) and switch open-close operations(writing). Further, for the conflicting write requests across partitions, a parent-child task queue is introduced. In the stress test of the on-line system of a provincial company with more than 10 million power grid equipment, response time less than 0.2 seconds is observed, under the load of more than 400 topology analysis requests per second per server.","PeriodicalId":399084,"journal":{"name":"2019 IEEE 14th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124847619","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 : 2019-11-01DOI: 10.1109/iske47853.2019.9170332
Kuo Pang, Ning Kang, Siqi Chen, Hongliang Zheng, L. Zou
With the coming of the era of big data, the recommendation system has become the most important way for all industries to obtain more effective information. Aiming at the problem of recommendation fuzzy explanation in recommendation system, we propose a collaborative filtering algorithm (CF) based on linguistic concept lattice with fuzzy object. Specifically, we introduce two operators between objects and linguistic concepts, and discuss their properties. Based on that, we construct a linguistic concept lattice with fuzzy object. Furthermore, considering the similarity between objects in the linguistic formal context with fuzzy object, we also propose an aggregation operator between linguistic concepts to obtain more accurate recommendation results for users. Finally, a practical example which concerned the movie recommendation is given to intuitively illustrate the applicability and effectiveness of the algorithm.
{"title":"Collaborative Filtering Recommendation Algorithm Based on Linguistic Concept Lattice with Fuzzy Object","authors":"Kuo Pang, Ning Kang, Siqi Chen, Hongliang Zheng, L. Zou","doi":"10.1109/iske47853.2019.9170332","DOIUrl":"https://doi.org/10.1109/iske47853.2019.9170332","url":null,"abstract":"With the coming of the era of big data, the recommendation system has become the most important way for all industries to obtain more effective information. Aiming at the problem of recommendation fuzzy explanation in recommendation system, we propose a collaborative filtering algorithm (CF) based on linguistic concept lattice with fuzzy object. Specifically, we introduce two operators between objects and linguistic concepts, and discuss their properties. Based on that, we construct a linguistic concept lattice with fuzzy object. Furthermore, considering the similarity between objects in the linguistic formal context with fuzzy object, we also propose an aggregation operator between linguistic concepts to obtain more accurate recommendation results for users. Finally, a practical example which concerned the movie recommendation is given to intuitively illustrate the applicability and effectiveness of the algorithm.","PeriodicalId":399084,"journal":{"name":"2019 IEEE 14th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128326171","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 : 2019-11-01DOI: 10.1109/ISKE47853.2019.9170283
Duo Zhao, Xiaying Zhang
Cross-Entropy (CE) optimization algorithm, whose characteristics are accurate and robust, has attracted widespread academic attention in recent years. A major drawback of CE algorithm is that it tends to be trapped in local optima. An advanced elite chaotic multi-objective cross entropy (ECCE) algorithm is proposed to enhance the search capability of CE algorithm confronting complex multimodal functions. Compared with the original algorithm, ECCE algorithm selects an elite individual to execute chaotic local search strategy. In the initial stage of algorithm, chaotic local search could explore search space to avoid premature convergence, it could also narrow search region in final stage to accurately locate optimal solution. The ECCE algorithm has been validated by standard test functions, and simulation results show that ECCE algorithm has certain advantages in optimizing multi-peak functions.
{"title":"A Multi-Objective Cross Entropy Algorithm Based on Elite Chaotic Local Search","authors":"Duo Zhao, Xiaying Zhang","doi":"10.1109/ISKE47853.2019.9170283","DOIUrl":"https://doi.org/10.1109/ISKE47853.2019.9170283","url":null,"abstract":"Cross-Entropy (CE) optimization algorithm, whose characteristics are accurate and robust, has attracted widespread academic attention in recent years. A major drawback of CE algorithm is that it tends to be trapped in local optima. An advanced elite chaotic multi-objective cross entropy (ECCE) algorithm is proposed to enhance the search capability of CE algorithm confronting complex multimodal functions. Compared with the original algorithm, ECCE algorithm selects an elite individual to execute chaotic local search strategy. In the initial stage of algorithm, chaotic local search could explore search space to avoid premature convergence, it could also narrow search region in final stage to accurately locate optimal solution. The ECCE algorithm has been validated by standard test functions, and simulation results show that ECCE algorithm has certain advantages in optimizing multi-peak functions.","PeriodicalId":399084,"journal":{"name":"2019 IEEE 14th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128417760","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 : 2019-11-01DOI: 10.1109/ISKE47853.2019.9170376
Yuxuan Wang, Bo Yu, Hui Shi, Xinyu He, Yonggong Ren
Biomedical event extraction is an important and challenging task in Information Extraction, which plays an important role for medicine research and disease prevention. Trigger identification has attracted much attention as the prerequisite step in biomedical event extraction. To skip the manual complex feature extraction, we propose a trigger identification method based on Bidirectional Long Short Term Memory (BLSTM) neural network. To obtain more semantic and syntactic information, we train dependency-based word embeddings to represent words, and add sentence embeddings to enrich sentence-level features. In addition, the attention mechanism is integrated to capture the most important semantic information in the sentence. The experimental results on the multi-level event extraction (MLEE) corpus show that the proposed method outperforms the state-of-the-art systems, achieving an F-score of 79.96%.
生物医学事件提取是信息提取领域的一项重要而富有挑战性的任务，在医学研究和疾病预防中发挥着重要作用。触发器识别作为生物医学事件提取的前提步骤，受到了广泛的关注。为了跳过人工复杂特征提取，提出了一种基于双向长短期记忆(Bidirectional Long - Short Term Memory, BLSTM)神经网络的触发器识别方法。为了获得更多的语义和句法信息，我们训练了基于依赖的词嵌入来表示单词，并增加了句子嵌入来丰富句子级特征。此外，还集成了注意机制，以捕获句子中最重要的语义信息。在多层事件提取(MLEE)语料库上的实验结果表明，该方法优于现有系统，f值达到79.96%。
{"title":"The Attention Based BLSTM Model Integrating Sentence Embeddings for Biomedical Event Trigger Identification","authors":"Yuxuan Wang, Bo Yu, Hui Shi, Xinyu He, Yonggong Ren","doi":"10.1109/ISKE47853.2019.9170376","DOIUrl":"https://doi.org/10.1109/ISKE47853.2019.9170376","url":null,"abstract":"Biomedical event extraction is an important and challenging task in Information Extraction, which plays an important role for medicine research and disease prevention. Trigger identification has attracted much attention as the prerequisite step in biomedical event extraction. To skip the manual complex feature extraction, we propose a trigger identification method based on Bidirectional Long Short Term Memory (BLSTM) neural network. To obtain more semantic and syntactic information, we train dependency-based word embeddings to represent words, and add sentence embeddings to enrich sentence-level features. In addition, the attention mechanism is integrated to capture the most important semantic information in the sentence. The experimental results on the multi-level event extraction (MLEE) corpus show that the proposed method outperforms the state-of-the-art systems, achieving an F-score of 79.96%.","PeriodicalId":399084,"journal":{"name":"2019 IEEE 14th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126904325","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 : 2019-11-01DOI: 10.1109/ISKE47853.2019.9170385
Xiaohong Liu, Xianyi Zeng
Since the 1980s, human beings have gradually created a knowledge economy with knowledge innovation as the main productive force. Knowledge employees are the most important resources in the era of knowledge economy. Because knowledge employee have more knowledge and have the ability to create new knowledge, they have strong liquidity, which can cause their value to be lost. On summarizing the concept and origin of knowledge employee, value-loss feature and its causes of knowledge employees is discussed. Based on the concept of utility in economics, the utility model of knowledge employees is analyzed, and the equilibrium model of vulnerability of knowledge employees is presented based on utility analysis in this paper
{"title":"An Evaluation of Value-Loss of Knowledge Employees","authors":"Xiaohong Liu, Xianyi Zeng","doi":"10.1109/ISKE47853.2019.9170385","DOIUrl":"https://doi.org/10.1109/ISKE47853.2019.9170385","url":null,"abstract":"Since the 1980s, human beings have gradually created a knowledge economy with knowledge innovation as the main productive force. Knowledge employees are the most important resources in the era of knowledge economy. Because knowledge employee have more knowledge and have the ability to create new knowledge, they have strong liquidity, which can cause their value to be lost. On summarizing the concept and origin of knowledge employee, value-loss feature and its causes of knowledge employees is discussed. Based on the concept of utility in economics, the utility model of knowledge employees is analyzed, and the equilibrium model of vulnerability of knowledge employees is presented based on utility analysis in this paper","PeriodicalId":399084,"journal":{"name":"2019 IEEE 14th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130663538","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 : 2019-11-01DOI: 10.1109/ISKE47853.2019.9170207
Yongsheng Liu, Wenyu Chen, S. H. Mahmud, Hong Qu, Kebin Miao, Feng Wei, Ziliang Zhang
The abundance of image-level labels and the lack of large scale bounding boxes detailed annotations promotes the expansion of Weakly-Supervised techniques for Object Detection (WSOD). In this work, we propose a novel mutual constraint learning for convolutional neural networks applied to detect bounding boxes only with global image-level supervision. The essence of our architecture is two new differentiable modules, Determination Network, and Parameterised Spatial Division, which explicitly allows the spatial division of the feature map within the network. These learnable modules give neural networks the ability to constructively generate shadow activation maps, dependent on the class activation maps. To demonstrate the effectiveness of our model for WSOD, we conduct extensive experiments on the multi-MNIST dataset. Experimental results show that mutual constraint learning can (i) help improve the accuracy of multi-category tasks, (ii) implement in an end-to-end way only with the image-level annotations, and (iii) output accurate bounding box labels, thereby achieving object detection.
{"title":"Mutual Constraint Learning for Weakly Supervised Object Detection","authors":"Yongsheng Liu, Wenyu Chen, S. H. Mahmud, Hong Qu, Kebin Miao, Feng Wei, Ziliang Zhang","doi":"10.1109/ISKE47853.2019.9170207","DOIUrl":"https://doi.org/10.1109/ISKE47853.2019.9170207","url":null,"abstract":"The abundance of image-level labels and the lack of large scale bounding boxes detailed annotations promotes the expansion of Weakly-Supervised techniques for Object Detection (WSOD). In this work, we propose a novel mutual constraint learning for convolutional neural networks applied to detect bounding boxes only with global image-level supervision. The essence of our architecture is two new differentiable modules, Determination Network, and Parameterised Spatial Division, which explicitly allows the spatial division of the feature map within the network. These learnable modules give neural networks the ability to constructively generate shadow activation maps, dependent on the class activation maps. To demonstrate the effectiveness of our model for WSOD, we conduct extensive experiments on the multi-MNIST dataset. Experimental results show that mutual constraint learning can (i) help improve the accuracy of multi-category tasks, (ii) implement in an end-to-end way only with the image-level annotations, and (iii) output accurate bounding box labels, thereby achieving object detection.","PeriodicalId":399084,"journal":{"name":"2019 IEEE 14th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130671984","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 : 2019-11-01DOI: 10.1109/ISKE47853.2019.9170408
Wenzhe Jin, Weihua Wu, Jialin Shi, Ming Gao
O2O instant delivery is a key success factor for internet-driven traditional business revolution such as 020 fast food delivery and fetch and carry services. With its rapid development, it has to face a variety of instant delivery scheduling problems. These have become a major bottleneck for the instant delivery platforms and it’s decision makers. The traditional method based on human experience has been no longer suitable for the current scenarios with large-scale orders and randomness. Therefore, we define the scheduling problem and scheduling strategies of instant orders delivery based on the real-world investigation, and verify the feasibility of order scheduling strategies by establishing an integrated simulation platform including traffic, orders, customers and merchants based on the DRL’s sumo simulation library and the comparison of the three scheduling strategies are used to select the optimal strategy. We simply designed these three scheduling strategies: (1) RD: The platform randomly assigns the order to the rider. (2) SP-D: The generated order is dispatched in real time to the rider closest to the order. (3) BPD: After all the orders generated in a fixed period of time are put together, all the orders are dispatched to several riders. It was found that the performance of BP-D was better in the comparison of SP-D and BP-D. And we found that when it comes to dealing with large-scale orders, the BP-D can most effectively dispatch passengers, maximizing the benefits of merchants, passengers and platforms.
{"title":"Simulation Based Scheduling Strategies Comparison of O2O Instant Delivery System","authors":"Wenzhe Jin, Weihua Wu, Jialin Shi, Ming Gao","doi":"10.1109/ISKE47853.2019.9170408","DOIUrl":"https://doi.org/10.1109/ISKE47853.2019.9170408","url":null,"abstract":"O2O instant delivery is a key success factor for internet-driven traditional business revolution such as 020 fast food delivery and fetch and carry services. With its rapid development, it has to face a variety of instant delivery scheduling problems. These have become a major bottleneck for the instant delivery platforms and it’s decision makers. The traditional method based on human experience has been no longer suitable for the current scenarios with large-scale orders and randomness. Therefore, we define the scheduling problem and scheduling strategies of instant orders delivery based on the real-world investigation, and verify the feasibility of order scheduling strategies by establishing an integrated simulation platform including traffic, orders, customers and merchants based on the DRL’s sumo simulation library and the comparison of the three scheduling strategies are used to select the optimal strategy. We simply designed these three scheduling strategies: (1) RD: The platform randomly assigns the order to the rider. (2) SP-D: The generated order is dispatched in real time to the rider closest to the order. (3) BPD: After all the orders generated in a fixed period of time are put together, all the orders are dispatched to several riders. It was found that the performance of BP-D was better in the comparison of SP-D and BP-D. And we found that when it comes to dealing with large-scale orders, the BP-D can most effectively dispatch passengers, maximizing the benefits of merchants, passengers and platforms.","PeriodicalId":399084,"journal":{"name":"2019 IEEE 14th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130839223","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 : 2019-11-01DOI: 10.1109/ISKE47853.2019.9170388
J. C. López, D. A. G. Chavira
In this article, we offer a novel multicriteria decision analysis method for the segmentation of the market. The proposed method combines the analysis of preferences of the customer and the application of decision aiding on the segmentation problem. To explore the preferences of each customer in a strong way, the method applies the aggregation-disaggregation paradigm and a genetic algorithm to derive multiple sets of preference parameters of the ELECTRE III method compatible with the preference information supplied by each customer. Next, the preferences of each customer are characterized by the dispersion of potential rankings of products by applying the derived valued outranking relations. A novel metric is used to quantify the similitude among preferences of diverse customers, and a procedure of clustering is established to complete the segmentation of the market.
{"title":"Multicriteria Market Segmentation: An Outranking Approach","authors":"J. C. López, D. A. G. Chavira","doi":"10.1109/ISKE47853.2019.9170388","DOIUrl":"https://doi.org/10.1109/ISKE47853.2019.9170388","url":null,"abstract":"In this article, we offer a novel multicriteria decision analysis method for the segmentation of the market. The proposed method combines the analysis of preferences of the customer and the application of decision aiding on the segmentation problem. To explore the preferences of each customer in a strong way, the method applies the aggregation-disaggregation paradigm and a genetic algorithm to derive multiple sets of preference parameters of the ELECTRE III method compatible with the preference information supplied by each customer. Next, the preferences of each customer are characterized by the dispersion of potential rankings of products by applying the derived valued outranking relations. A novel metric is used to quantify the similitude among preferences of diverse customers, and a procedure of clustering is established to complete the segmentation of the market.","PeriodicalId":399084,"journal":{"name":"2019 IEEE 14th International Conference on Intelligent Systems and Knowledge Engineering (ISKE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131340992","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：