The purpose of this paper is to study personalized needs of e-commerce website category hierarchy based on users' mental models by means of Multidimensional Scaling and Web Usage Mining. The users' browsing category paths in an e-commerce website is extracted based on the Web Usage Mining, and the Multidimensional Scaling was used to probe the structure and composition of the users' mental models of website category hierarchy based on their browsing category paths, at last, users' personalized needs can be identified. Three million web log data records were collected for experimental study. The experimental results show the proposed method is efficient to discover users' personalized needs of expected category hierarchy based on large scale web log data automatically and efficiently.
{"title":"Personalized Recommendation of E-Commerce Website Category Hierarchy Based on Web Usage Mining and Multidimensional Scaling","authors":"Peng Wu, Jiamin Wang, Daqing He","doi":"10.1109/ICEBE.2015.23","DOIUrl":"https://doi.org/10.1109/ICEBE.2015.23","url":null,"abstract":"The purpose of this paper is to study personalized needs of e-commerce website category hierarchy based on users' mental models by means of Multidimensional Scaling and Web Usage Mining. The users' browsing category paths in an e-commerce website is extracted based on the Web Usage Mining, and the Multidimensional Scaling was used to probe the structure and composition of the users' mental models of website category hierarchy based on their browsing category paths, at last, users' personalized needs can be identified. Three million web log data records were collected for experimental study. The experimental results show the proposed method is efficient to discover users' personalized needs of expected category hierarchy based on large scale web log data automatically and efficiently.","PeriodicalId":153535,"journal":{"name":"2015 IEEE 12th International Conference on e-Business Engineering","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126125015","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
With rapid development of E-commerce platforms, automated review sentiment analysis for commodities becomes a research focus, with main purpose to extract potential information within reviews for decision making of consumers. Traditional methods have made some progress on document level sentiment analysis, but with tremendous increasing of data scale, how to process high dimension of data fast and effectively becomes the largest limitation. In this paper, we import deep neural network which is appropriate for high dimension data analysis, and propose a framework of sentiment analysis based on deep learning. Experiments on different data scale and different domains show that the proposed method can solve high dimensional problem with good performance.
{"title":"Review Sentiment Analysis Based on Deep Learning","authors":"Zhongkai Hu, Jianqing Hu, Weifeng Ding, Xiaolin Zheng","doi":"10.1109/ICEBE.2015.24","DOIUrl":"https://doi.org/10.1109/ICEBE.2015.24","url":null,"abstract":"With rapid development of E-commerce platforms, automated review sentiment analysis for commodities becomes a research focus, with main purpose to extract potential information within reviews for decision making of consumers. Traditional methods have made some progress on document level sentiment analysis, but with tremendous increasing of data scale, how to process high dimension of data fast and effectively becomes the largest limitation. In this paper, we import deep neural network which is appropriate for high dimension data analysis, and propose a framework of sentiment analysis based on deep learning. Experiments on different data scale and different domains show that the proposed method can solve high dimensional problem with good performance.","PeriodicalId":153535,"journal":{"name":"2015 IEEE 12th International Conference on e-Business Engineering","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130244820","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 current Internet environment, security incidents occur frequently and the potential safety problems are highlighted. In this case, we propose "Internet trusted identity authentication system". The system changes the previous conventional mode of identity authentication, changes the mode of username/password to the relationship bound between net ID and password rule and replaces the traditional dead-password mode with the mode of dynamic password and password rule. At the same time, we should establish the repository of trusted identity to store and manage the net ID and password rule. With the tools of USBKEY, SMS and digital certificates, the system can achieve the function of implementing the real-name network in the Internet, making remembering passwords more convenience, simplifying users' operations online and strengthening the security of personal identification information online, which can improve the safety, reliability and efficiency of the Internet, on the premise that it won't change the Internet functions or the operation habits.
{"title":"Study on the Model and Algorithm of Internet Trusted Identity Authentication System","authors":"Yao-Ching Liu, Y. Chai, Yi Liu","doi":"10.1109/ICEBE.2015.76","DOIUrl":"https://doi.org/10.1109/ICEBE.2015.76","url":null,"abstract":"In the current Internet environment, security incidents occur frequently and the potential safety problems are highlighted. In this case, we propose \"Internet trusted identity authentication system\". The system changes the previous conventional mode of identity authentication, changes the mode of username/password to the relationship bound between net ID and password rule and replaces the traditional dead-password mode with the mode of dynamic password and password rule. At the same time, we should establish the repository of trusted identity to store and manage the net ID and password rule. With the tools of USBKEY, SMS and digital certificates, the system can achieve the function of implementing the real-name network in the Internet, making remembering passwords more convenience, simplifying users' operations online and strengthening the security of personal identification information online, which can improve the safety, reliability and efficiency of the Internet, on the premise that it won't change the Internet functions or the operation habits.","PeriodicalId":153535,"journal":{"name":"2015 IEEE 12th International Conference on e-Business Engineering","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132420804","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}
Sara Capecchi, Mauro Giraudo, Claudia Negro, P. Pisano
In this paper we analyse the use of an extension of tools for service design and business conceptual modelling toward reputation system design for the realisation of Schooladvise a platform for educational services reputation management. The work described in the present paper originates from the intuition that business process modelling, service design and reputation system modelling share many needs, aims and problems: to represent a variety of entities involved in the related domains, to improve services/business in order to meet the users need, moreover they all involve stakeholders with very different background and skills, a problem known as socio-technical mismatch. The objective and the contribution of the analysis presented in this paper is to gather feedback on the proposed modelling language from the end-users (requirements engineers and domain experts) while applying it to a complex case study.
{"title":"Schooladvise: Designing a Reputation System for Educational Services through Service Design and Business Modeling","authors":"Sara Capecchi, Mauro Giraudo, Claudia Negro, P. Pisano","doi":"10.1109/ICEBE.2015.52","DOIUrl":"https://doi.org/10.1109/ICEBE.2015.52","url":null,"abstract":"In this paper we analyse the use of an extension of tools for service design and business conceptual modelling toward reputation system design for the realisation of Schooladvise a platform for educational services reputation management. The work described in the present paper originates from the intuition that business process modelling, service design and reputation system modelling share many needs, aims and problems: to represent a variety of entities involved in the related domains, to improve services/business in order to meet the users need, moreover they all involve stakeholders with very different background and skills, a problem known as socio-technical mismatch. The objective and the contribution of the analysis presented in this paper is to gather feedback on the proposed modelling language from the end-users (requirements engineers and domain experts) while applying it to a complex case study.","PeriodicalId":153535,"journal":{"name":"2015 IEEE 12th International Conference on e-Business Engineering","volume":"90 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132568352","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}
Hao-Ting Pai, Fan Wu, P. Hsueh, Grace Lin, Ya-Hui Chan
Recently, enterprises and governments invested aggressively in big data analytics because it is truly representative of popular opinion based on millions of people. Despite bringing new opportunities, big data encounters the challenges such as extremely large number of observations (e.g., Millions of transactions), high dimensionality (e.g., Thousands of items), and immediate response. Taking big data into consideration, the conventional association analysis is frustrated by the extraction of patterns information. Specifically, the computational complexity of frequent item sets mining increases exponentially by the number of items, which has been proven to be an NP-Complete problem. Although many studies used a pruning-patterns strategy to reduce the complexity, it probably distorts the shape of data and incurs inaccurate result. In this paper, we introduce relative patterns discovery (named RPD) that explores the same patterns between each two observations. To show that RPD is a pragmatic solution toward big data analytics, we design a scalable outlier detection method (named SOD) based on the concept of RPD. Particularly, SOD can score the anomaly without enumerate all the relative patterns. The empirical investigations, conducted with various real-world datasets, demonstrate that SOD performs well even in the environment of large number of observations and high dimensionality.
{"title":"Relative Patterns Discovery toward Big Data Analytics","authors":"Hao-Ting Pai, Fan Wu, P. Hsueh, Grace Lin, Ya-Hui Chan","doi":"10.1109/ICEBE.2015.74","DOIUrl":"https://doi.org/10.1109/ICEBE.2015.74","url":null,"abstract":"Recently, enterprises and governments invested aggressively in big data analytics because it is truly representative of popular opinion based on millions of people. Despite bringing new opportunities, big data encounters the challenges such as extremely large number of observations (e.g., Millions of transactions), high dimensionality (e.g., Thousands of items), and immediate response. Taking big data into consideration, the conventional association analysis is frustrated by the extraction of patterns information. Specifically, the computational complexity of frequent item sets mining increases exponentially by the number of items, which has been proven to be an NP-Complete problem. Although many studies used a pruning-patterns strategy to reduce the complexity, it probably distorts the shape of data and incurs inaccurate result. In this paper, we introduce relative patterns discovery (named RPD) that explores the same patterns between each two observations. To show that RPD is a pragmatic solution toward big data analytics, we design a scalable outlier detection method (named SOD) based on the concept of RPD. Particularly, SOD can score the anomaly without enumerate all the relative patterns. The empirical investigations, conducted with various real-world datasets, demonstrate that SOD performs well even in the environment of large number of observations and high dimensionality.","PeriodicalId":153535,"journal":{"name":"2015 IEEE 12th International Conference on e-Business Engineering","volume":"207 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121616128","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}
Shuai Luan, Hongming Cai, Fenglin Bu, Lihong Jiang
Huge amounts of spatial documents have been generated every day in IoT (internet of things) applications. Both spatial and textual or other information of complicated structure should be taken into account when providing services. Therefore, it is crucial to store and process these data efficiently in IoT applications. In this paper, we propose a flexible four-layer framework for spatial data towards IoT applications, which decouples the spatial and content information and organizes the whole space with a pyramid structure. The four layers of the framework are geography node, application node, logical node and storage node that are related with the accurate spatial information, content information with rich semantic, flexible space organization and efficient storage with hybrid database respectively. Among them, logical node is the center which associates the other nodes, and provides an efficient query processing method using the pyramid structure. We apply the four-layer framework in the Smarter City Charging Pile System and construct the four kinds of nodes to realize the function that recommends the best matching charging piles for electric vehicle drivers taking both distance and pile state into consideration. The compare discussion shows that our framework has high flexibility, supports rich semantic and is more acceptable in practical IoT applications.
{"title":"A Four-Layer Flexible Spatial Data Framework towards IoT Application","authors":"Shuai Luan, Hongming Cai, Fenglin Bu, Lihong Jiang","doi":"10.1109/ICEBE.2015.64","DOIUrl":"https://doi.org/10.1109/ICEBE.2015.64","url":null,"abstract":"Huge amounts of spatial documents have been generated every day in IoT (internet of things) applications. Both spatial and textual or other information of complicated structure should be taken into account when providing services. Therefore, it is crucial to store and process these data efficiently in IoT applications. In this paper, we propose a flexible four-layer framework for spatial data towards IoT applications, which decouples the spatial and content information and organizes the whole space with a pyramid structure. The four layers of the framework are geography node, application node, logical node and storage node that are related with the accurate spatial information, content information with rich semantic, flexible space organization and efficient storage with hybrid database respectively. Among them, logical node is the center which associates the other nodes, and provides an efficient query processing method using the pyramid structure. We apply the four-layer framework in the Smarter City Charging Pile System and construct the four kinds of nodes to realize the function that recommends the best matching charging piles for electric vehicle drivers taking both distance and pile state into consideration. The compare discussion shows that our framework has high flexibility, supports rich semantic and is more acceptable in practical IoT applications.","PeriodicalId":153535,"journal":{"name":"2015 IEEE 12th International Conference on e-Business Engineering","volume":"118 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114370315","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper presents a quantitative e-business system design method. At first we put forward the requirements for e-business quantitatively. Then we put forward a quantitative e-business design model for e-business system. In this paper we try to find digital values in the process of software architecture design. Our target is to design software architectures more objectively, which makes the software architectural design process become an effective quantitative process. Finally, we present an example for e-business system in this paper.
{"title":"Quantitative E-Business System Design Method","authors":"Lixin Zhou","doi":"10.1109/ICEBE.2015.80","DOIUrl":"https://doi.org/10.1109/ICEBE.2015.80","url":null,"abstract":"This paper presents a quantitative e-business system design method. At first we put forward the requirements for e-business quantitatively. Then we put forward a quantitative e-business design model for e-business system. In this paper we try to find digital values in the process of software architecture design. Our target is to design software architectures more objectively, which makes the software architectural design process become an effective quantitative process. Finally, we present an example for e-business system in this paper.","PeriodicalId":153535,"journal":{"name":"2015 IEEE 12th International Conference on e-Business Engineering","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115048978","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}
Traditionally, direct weighting method is adopted in the calculation of price index in a region. Under the online shopping environment, there are significant differences in consumer spending in different E-commerce enterprises, because the business scope, the main commodities, and the target consumers vary a lot in different enterprises. If the direct weighting method is adopted, the price index of an E-commerce enterprise with lower price increase of all commodities may be higher than the one with higher price increase. The solution is there should be diverse weights of commodities in different enterprises. This paper proposes the Hedonic method towards higher-level compilation of online CPI: elementary price index is regarded as a function of enterprise property and commodity property, and price index by category is calculated using least square method. In addition, the linear form and exponential form of Hedonic method are provided. At last, the Hedonic method is tested by calculating the chain and fixed base price index with real data, and the result shows that not only the testing results but also the price index of the linear form and the exponential form are very similar.
{"title":"The Hedonic Method towards Higher-Level Compilation of Online CPI","authors":"Yadong Huang, Y. Chai, Yi Liu, H. Chu","doi":"10.1109/ICEBE.2015.37","DOIUrl":"https://doi.org/10.1109/ICEBE.2015.37","url":null,"abstract":"Traditionally, direct weighting method is adopted in the calculation of price index in a region. Under the online shopping environment, there are significant differences in consumer spending in different E-commerce enterprises, because the business scope, the main commodities, and the target consumers vary a lot in different enterprises. If the direct weighting method is adopted, the price index of an E-commerce enterprise with lower price increase of all commodities may be higher than the one with higher price increase. The solution is there should be diverse weights of commodities in different enterprises. This paper proposes the Hedonic method towards higher-level compilation of online CPI: elementary price index is regarded as a function of enterprise property and commodity property, and price index by category is calculated using least square method. In addition, the linear form and exponential form of Hedonic method are provided. At last, the Hedonic method is tested by calculating the chain and fixed base price index with real data, and the result shows that not only the testing results but also the price index of the linear form and the exponential form are very similar.","PeriodicalId":153535,"journal":{"name":"2015 IEEE 12th International Conference on e-Business Engineering","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115215724","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}
A novel method to measure the graph similarity is proposed, where the labels, in-degrees, and out-degrees of the vertices in the graph are comprehensively considered in order to conquer the high complexity and information loss in the measurement process. In this proposal, the graph is decomposed into a generalized tree and its similarity can be measured by computing the cost of the node sequence triples in the generalized tree. The proposed method decreases the computational complexity to O(n2) in comparison with the other available graph similarity methods. An example is provided to illustrate that the method is feasible and effective.
{"title":"A novel method to measure graph similarity","authors":"Xu Wang, Jihong Ouyang","doi":"10.1109/ICEBE.2015.38","DOIUrl":"https://doi.org/10.1109/ICEBE.2015.38","url":null,"abstract":"A novel method to measure the graph similarity is proposed, where the labels, in-degrees, and out-degrees of the vertices in the graph are comprehensively considered in order to conquer the high complexity and information loss in the measurement process. In this proposal, the graph is decomposed into a generalized tree and its similarity can be measured by computing the cost of the node sequence triples in the generalized tree. The proposed method decreases the computational complexity to O(n2) in comparison with the other available graph similarity methods. An example is provided to illustrate that the method is feasible and effective.","PeriodicalId":153535,"journal":{"name":"2015 IEEE 12th International Conference on e-Business Engineering","volume":"32 3","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121015061","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}
Exact document interpretation is very important to semantic document exchange. An essential issue for document interpretation is to maintain semantic consistency of the exchanged documents between any two autonomous business communities, where the document sender and receiver have no misunderstanding in using the exchanged documents. Existing approaches to resolving this issue adopt document standards or apply business ontologies. While these approaches are effective in certain degree, the issue of limited flexibility and evolvement in using standards and the lack of accuracy in using ontologies to mediate document elements must be explored and resolved. This paper proposes a multi-viewed tabular document (Tabdoc) approach to achieving consistent document interpretation between document writers and readers. In this approach, any document is a table, which is seen as a tree to represent document structure. Concepts, layout and grammatical relations of a same document are separated from the document structure and form different views in Tabdoc model.
{"title":"A Multi-viewed Document Representation for Semantic Document Exchange","authors":"Shuo Yang, J. Guo","doi":"10.1109/ICEBE.2015.34","DOIUrl":"https://doi.org/10.1109/ICEBE.2015.34","url":null,"abstract":"Exact document interpretation is very important to semantic document exchange. An essential issue for document interpretation is to maintain semantic consistency of the exchanged documents between any two autonomous business communities, where the document sender and receiver have no misunderstanding in using the exchanged documents. Existing approaches to resolving this issue adopt document standards or apply business ontologies. While these approaches are effective in certain degree, the issue of limited flexibility and evolvement in using standards and the lack of accuracy in using ontologies to mediate document elements must be explored and resolved. This paper proposes a multi-viewed tabular document (Tabdoc) approach to achieving consistent document interpretation between document writers and readers. In this approach, any document is a table, which is seen as a tree to represent document structure. Concepts, layout and grammatical relations of a same document are separated from the document structure and form different views in Tabdoc model.","PeriodicalId":153535,"journal":{"name":"2015 IEEE 12th International Conference on e-Business Engineering","volume":"2017 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-10-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134355419","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：