With the rapid development of big data,online education can use big data technology to achieve personalized and intelligent education as well as improve learning effect and user satisfaction.In this study,the users of...With the rapid development of big data,online education can use big data technology to achieve personalized and intelligent education as well as improve learning effect and user satisfaction.In this study,the users of The Open University of China online education platform were taken as the research object,their user behavior data was collected,cleaned,and analyzed with text mining.The RFM model and the improved K-Means algorithm were used to construct the user portrait of the platform group and the needs and preferences of different types of the users were analyzded.Chinese word segmentation was used to show the key words of different types of users and the word cloud of their using frequency.The focus of different user groups was determined to facilitate for the follow-up course recommendation and precision marketing.Experimental results showed that the improved K-Means algorithm can well depict the behavior of group users.The index of silhouette score was improved to 0.811 by the improved K-Means algorithm,from random uncertainty to a fixed value,which can effectively solve the problem of inconsistent results caused by outlier sample points.展开更多
Hotspot topic trends can be captured by analyzing user attributes and historical behavior in social network. In this paper, we propose a user participation behavior prediction model for social hotspots, based on user ...Hotspot topic trends can be captured by analyzing user attributes and historical behavior in social network. In this paper, we propose a user participation behavior prediction model for social hotspots, based on user behavior and relationship data, to predict user participation behavior and topic development trends. Firstly, for the complex factors of user behavior, three dynamic influence factor functions are defined, including individual, peer and community influence. These functions take timeliness into account using a time discretization method. Secondly, to determine laws of individual behavior and group behavior within a social topic, a hotspot user participation behavior prediction model is proposed and associated with the basic concepts of randora field and Markov property in information diffusion. The experimental results show that the model can not only dynamically predict the individual behavior, but also grasp the development trends of topics.展开更多
Microblogs have become an important platform for people to publish,transform information and acquire knowledge.This paper focuses on the problem of discovering user interest in microblogs.In this paper,we propose a to...Microblogs have become an important platform for people to publish,transform information and acquire knowledge.This paper focuses on the problem of discovering user interest in microblogs.In this paper,we propose a topic mining model based on Latent Dirichlet Allocation(LDA) named user-topic model.For each user,the interests are divided into two parts by different ways to generate the microblogs:original interest and retweet interest.We represent a Gibbs sampling implementation for inference the parameters of our model,and discover not only user's original interest,but also retweet interest.Then we combine original interest and retweet interest to compute interest words for users.Experiments on a dataset of Sina microblogs demonstrate that our model is able to discover user interest effectively and outperforms existing topic models in this task.And we find that original interest and retweet interest are similar and the topics of interest contain user labels.The interest words discovered by our model reflect user labels,but range is much broader.展开更多
Collaborative Filtering(CF) is a leading approach to build recommender systems which has gained considerable development and popularity. A predominant approach to CF is rating prediction recommender algorithm, aiming ...Collaborative Filtering(CF) is a leading approach to build recommender systems which has gained considerable development and popularity. A predominant approach to CF is rating prediction recommender algorithm, aiming to predict a user's rating for those items which were not rated yet by the user. However, with the increasing number of items and users, thedata is sparse.It is difficult to detectlatent closely relation among the items or users for predicting the user behaviors. In this paper,we enhance the rating prediction approach leading to substantial improvement of prediction accuracy by categorizing according to the genres of movies. Then the probabilities that users are interested in the genres are computed to integrate the prediction of each genre cluster. A novel probabilistic approach based on the sentiment analysis of the user reviews is also proposed to give intuitional explanations of why an item is recommended.To test the novel recommendation approach, a new corpus of user reviews on movies obtained from the Internet Movies Database(IMDB) has been generated. Experimental results show that the proposed framework is effective and achieves a better prediction performance.展开更多
Service modeling offers fundamental guidance to the construction and operation of mobile telecommunication networks. As the domestic LTE networks have been deployed massively, the refined LTE services model needs to b...Service modeling offers fundamental guidance to the construction and operation of mobile telecommunication networks. As the domestic LTE networks have been deployed massively, the refined LTE services model needs to be established urgently. In this paper, we firstly extract characteristic parameters of services from statistical data in 3G networks, especially in time, space and user dimension. Secondly, the development trends of LTE services are analyzed. And the refined LTE service model is established. Finally, prediction results of LTE service development in China is given, which could provide effi cient support for networks' optimization and evolution.展开更多
In order to construct the trusted network and realize the trust of network behavior,a new multi-dimensional behavior measurement model based on prediction and control is presented.By using behavior predictive equation...In order to construct the trusted network and realize the trust of network behavior,a new multi-dimensional behavior measurement model based on prediction and control is presented.By using behavior predictive equation,individual similarity function,group similarity function,direct trust assessment function,and generalized predictive control,this model can guarantee the trust of an end user and users in its network.Compared with traditional measurement model,the model considers different characteristics of various networks.The trusted measurement policies established according to different network environments have better adaptability.By constructing trusted group,the threats to trusted group will be reduced greatly.Utilizing trusted group to restrict individuals in network can ensure the fault tolerance of trustworthiness of trusted individuals and group.The simulation shows that this scheme can support behavior measurement more efficiently than traditional ones and the model resists viruses and Trojans more efficiently than older ones.展开更多
With increasingly complex website structure and continuously advancing web technologies,accurate user clicks recognition from massive HTTP data,which is critical for web usage mining,becomes more difficult.In this pap...With increasingly complex website structure and continuously advancing web technologies,accurate user clicks recognition from massive HTTP data,which is critical for web usage mining,becomes more difficult.In this paper,we propose a dependency graph model to describe the relationships between web requests.Based on this model,we design and implement a heuristic parallel algorithm to distinguish user clicks with the assistance of cloud computing technology.We evaluate the proposed algorithm with real massive data.The size of the dataset collected from a mobile core network is 228.7GB.It covers more than three million users.The experiment results demonstrate that the proposed algorithm can achieve higher accuracy than previous methods.展开更多
With the increasing popularity of cloud computing, there is an increased de mand for cloud resources in cloud. It has be come even more urgent to find solutions to improve resource utilization. From the per spective o...With the increasing popularity of cloud computing, there is an increased de mand for cloud resources in cloud. It has be come even more urgent to find solutions to improve resource utilization. From the per spective of a cloud consumer, a cloud applica tion processes a large information flow in volving user actions that access resources, but little work has so far been devoted to research from the perspective of the interaction be tween the user and the cloud application. In this paper, we analyze the interaction in detail, and propose a general mathematical interac tion model to formulate the challenge pertain ing to storage resource allocation as an opti mization problem, focusing on minimizing both the user＇s cost and server＇s consumption. A potential response mechanism is then de signed based on the interaction model. Fur thermore, the proposed model is used to ex plore strategies when multiple users access the same file simultaneously. Additionally, an improved queuing system, namely M/ G~ oo queue with standby, is introduced. Finally, an evaluation is presented to verify the interac- tion model.展开更多
Context-aware plays a pivotal role in the wisdom network of information processing. Due to the limited resources, context conflict is inevitable in context-aware. User satisfaction is a good reflection for the wisdom ...Context-aware plays a pivotal role in the wisdom network of information processing. Due to the limited resources, context conflict is inevitable in context-aware. User satisfaction is a good reflection for the wisdom degree of the network. In this paper, considering the user satisfaction, we propose a novel context-aware conflict solution model based on set pair Analysis （SPA）. The model obtains the best server mode by the maximum satisfaction connection degree to solve the conflict in context-aware. The simulation analysis shows that the proposed method is effective and can solve many users competition for the same scene of conflict.展开更多
This paper focuses on how to improve aspect-level opinion mining for online customer reviews. We first propose a novel generative topic model, the Joint Aspect/Sentiment (JAS) model, to jointly extract aspects and asp...This paper focuses on how to improve aspect-level opinion mining for online customer reviews. We first propose a novel generative topic model, the Joint Aspect/Sentiment (JAS) model, to jointly extract aspects and aspect-dependent sentiment lexicons from online customer reviews. An aspect-dependent sentiment lexicon refers to the aspect-specific opinion words along with their aspect-aware sentiment polarities with respect to a specific aspect. We then apply the extracted aspectdependent sentiment lexicons to a series of aspect-level opinion mining tasks, including implicit aspect identification, aspect-based extractive opinion summarization, and aspect-level sentiment classification. Experimental results demonstrate the effectiveness of the JAS model in learning aspectdependent sentiment lexicons and the practical values of the extracted lexicons when applied to these practical tasks.展开更多
Online social networks have gradually permeated into every aspect of people's life.As a research hotspot in social network, user influence is of theoretical and practical significant for information transmission, ...Online social networks have gradually permeated into every aspect of people's life.As a research hotspot in social network, user influence is of theoretical and practical significant for information transmission, optimization and integration. A prominent application is a viral marketing campaign which aims to use a small number of targeted infl uence users to initiate cascades of infl uence that create a global increase in product adoption. In this paper, we analyze mainly evaluation methods of user infl uence based on IDM evaluation model, Page Rank evaluation model, use behavior model and some other popular influence evaluation models in currently social network. And then, we extract the core idea of these models to build our influence evaluation model from two aspects, relationship and activity. Finally, the proposed approach was validated on real world datasets,and the result of experiments shows that our method is both effective and stable.展开更多
文摘With the rapid development of big data,online education can use big data technology to achieve personalized and intelligent education as well as improve learning effect and user satisfaction.In this study,the users of The Open University of China online education platform were taken as the research object,their user behavior data was collected,cleaned,and analyzed with text mining.The RFM model and the improved K-Means algorithm were used to construct the user portrait of the platform group and the needs and preferences of different types of the users were analyzded.Chinese word segmentation was used to show the key words of different types of users and the word cloud of their using frequency.The focus of different user groups was determined to facilitate for the follow-up course recommendation and precision marketing.Experimental results showed that the improved K-Means algorithm can well depict the behavior of group users.The index of silhouette score was improved to 0.811 by the improved K-Means algorithm,from random uncertainty to a fixed value,which can effectively solve the problem of inconsistent results caused by outlier sample points.
基金supported by the National Key Basic Research Program(973 program)of China(No.2013CB329606)National Science Foundation of China(Grant No.61272400)+2 种基金Science and Technology Research Program of the Chongqing Municipal Education Committee(No.KJ1500425)Wen Feng Foundation of CQUPT(No.WF201403)Chongqing Graduate Research And Innovation Project(No.CYS14146)
文摘Hotspot topic trends can be captured by analyzing user attributes and historical behavior in social network. In this paper, we propose a user participation behavior prediction model for social hotspots, based on user behavior and relationship data, to predict user participation behavior and topic development trends. Firstly, for the complex factors of user behavior, three dynamic influence factor functions are defined, including individual, peer and community influence. These functions take timeliness into account using a time discretization method. Secondly, to determine laws of individual behavior and group behavior within a social topic, a hotspot user participation behavior prediction model is proposed and associated with the basic concepts of randora field and Markov property in information diffusion. The experimental results show that the model can not only dynamically predict the individual behavior, but also grasp the development trends of topics.
基金This work was supported by the National High Technology Research and Development Program of China（No. 2010AA012505, 2011AA010702, 2012AA01A401 and 2012AA01A402）, Chinese National Science Foundation （No. 60933005, 91124002,61303265）, National Technology Support Foundation （No. 2012BAH38B04） and National 242 Foundation （No. 2011A010）
文摘Microblogs have become an important platform for people to publish,transform information and acquire knowledge.This paper focuses on the problem of discovering user interest in microblogs.In this paper,we propose a topic mining model based on Latent Dirichlet Allocation(LDA) named user-topic model.For each user,the interests are divided into two parts by different ways to generate the microblogs:original interest and retweet interest.We represent a Gibbs sampling implementation for inference the parameters of our model,and discover not only user's original interest,but also retweet interest.Then we combine original interest and retweet interest to compute interest words for users.Experiments on a dataset of Sina microblogs demonstrate that our model is able to discover user interest effectively and outperforms existing topic models in this task.And we find that original interest and retweet interest are similar and the topics of interest contain user labels.The interest words discovered by our model reflect user labels,but range is much broader.
基金supported in part by National Science Foundation of China under Grants No.61303105 and 61402304the Humanity&Social Science general project of Ministry of Education under Grants No.14YJAZH046+2 种基金the Beijing Natural Science Foundation under Grants No.4154065the Beijing Educational Committee Science and Technology Development Planned under Grants No.KM201410028017Academic Degree Graduate Courses group projects
文摘Collaborative Filtering(CF) is a leading approach to build recommender systems which has gained considerable development and popularity. A predominant approach to CF is rating prediction recommender algorithm, aiming to predict a user's rating for those items which were not rated yet by the user. However, with the increasing number of items and users, thedata is sparse.It is difficult to detectlatent closely relation among the items or users for predicting the user behaviors. In this paper,we enhance the rating prediction approach leading to substantial improvement of prediction accuracy by categorizing according to the genres of movies. Then the probabilities that users are interested in the genres are computed to integrate the prediction of each genre cluster. A novel probabilistic approach based on the sentiment analysis of the user reviews is also proposed to give intuitional explanations of why an item is recommended.To test the novel recommendation approach, a new corpus of user reviews on movies obtained from the Internet Movies Database(IMDB) has been generated. Experimental results show that the proposed framework is effective and achieves a better prediction performance.
文摘Service modeling offers fundamental guidance to the construction and operation of mobile telecommunication networks. As the domestic LTE networks have been deployed massively, the refined LTE services model needs to be established urgently. In this paper, we firstly extract characteristic parameters of services from statistical data in 3G networks, especially in time, space and user dimension. Secondly, the development trends of LTE services are analyzed. And the refined LTE service model is established. Finally, prediction results of LTE service development in China is given, which could provide effi cient support for networks' optimization and evolution.
基金This work was supported by the National Basic Research Pro-gram of China under Crant No.2007CB311100 Funds of Key Lab of Fujlan Province University Network Security and Cryp- toll1009+3 种基金 the National Science Foundation for Young Scholars of China under Crant No.61001091 Beijing Nature Science Foundation under Crant No. 4122012 ＂Next-Generation Broad-band Wireless Mobile Communication Network＂ National Sci-ence and Technology Major Special Issue Funding under Grant No. 2012ZX03002003 Funding Program for Academic tturmn Resources Development in Institutions of Higher Learning Under the Jurisdiction of Beijing Municipality of Chi-na.
文摘In order to construct the trusted network and realize the trust of network behavior,a new multi-dimensional behavior measurement model based on prediction and control is presented.By using behavior predictive equation,individual similarity function,group similarity function,direct trust assessment function,and generalized predictive control,this model can guarantee the trust of an end user and users in its network.Compared with traditional measurement model,the model considers different characteristics of various networks.The trusted measurement policies established according to different network environments have better adaptability.By constructing trusted group,the threats to trusted group will be reduced greatly.Utilizing trusted group to restrict individuals in network can ensure the fault tolerance of trustworthiness of trusted individuals and group.The simulation shows that this scheme can support behavior measurement more efficiently than traditional ones and the model resists viruses and Trojans more efficiently than older ones.
基金supported in part by the Fundamental Research Funds for the Central Universities under Grant No.2013RC0114111 Project of China under Grant No.B08004
文摘With increasingly complex website structure and continuously advancing web technologies,accurate user clicks recognition from massive HTTP data,which is critical for web usage mining,becomes more difficult.In this paper,we propose a dependency graph model to describe the relationships between web requests.Based on this model,we design and implement a heuristic parallel algorithm to distinguish user clicks with the assistance of cloud computing technology.We evaluate the proposed algorithm with real massive data.The size of the dataset collected from a mobile core network is 228.7GB.It covers more than three million users.The experiment results demonstrate that the proposed algorithm can achieve higher accuracy than previous methods.
基金supported in part by the National Natural Science Foundation of China under Grant No. 61271199the Fundamental Research Funds in Beijing Jiaotong University under Grant No. W11JB00630
文摘With the increasing popularity of cloud computing, there is an increased de mand for cloud resources in cloud. It has be come even more urgent to find solutions to improve resource utilization. From the per spective of a cloud consumer, a cloud applica tion processes a large information flow in volving user actions that access resources, but little work has so far been devoted to research from the perspective of the interaction be tween the user and the cloud application. In this paper, we analyze the interaction in detail, and propose a general mathematical interac tion model to formulate the challenge pertain ing to storage resource allocation as an opti mization problem, focusing on minimizing both the user＇s cost and server＇s consumption. A potential response mechanism is then de signed based on the interaction model. Fur thermore, the proposed model is used to ex plore strategies when multiple users access the same file simultaneously. Additionally, an improved queuing system, namely M/ G~ oo queue with standby, is introduced. Finally, an evaluation is presented to verify the interac- tion model.
文摘Context-aware plays a pivotal role in the wisdom network of information processing. Due to the limited resources, context conflict is inevitable in context-aware. User satisfaction is a good reflection for the wisdom degree of the network. In this paper, considering the user satisfaction, we propose a novel context-aware conflict solution model based on set pair Analysis （SPA）. The model obtains the best server mode by the maximum satisfaction connection degree to solve the conflict in context-aware. The simulation analysis shows that the proposed method is effective and can solve many users competition for the same scene of conflict.
基金supported by National Natural Science Foundation of China under Grants No.61232010, No.60903139, No.60933005, No.61202215, No.61100083National 242 Project under Grant No.2011F65China Information Technology Security Evaluation Center Program under Grant No.Z1277
文摘This paper focuses on how to improve aspect-level opinion mining for online customer reviews. We first propose a novel generative topic model, the Joint Aspect/Sentiment (JAS) model, to jointly extract aspects and aspect-dependent sentiment lexicons from online customer reviews. An aspect-dependent sentiment lexicon refers to the aspect-specific opinion words along with their aspect-aware sentiment polarities with respect to a specific aspect. We then apply the extracted aspectdependent sentiment lexicons to a series of aspect-level opinion mining tasks, including implicit aspect identification, aspect-based extractive opinion summarization, and aspect-level sentiment classification. Experimental results demonstrate the effectiveness of the JAS model in learning aspectdependent sentiment lexicons and the practical values of the extracted lexicons when applied to these practical tasks.
基金supported by the Research Fund for the Doctoral Program(New Teachers)Ministry of Education of China under Grant No.20121103120032+2 种基金Humanity and Social Science Youth foundation of Ministry of Education of China under Grant No.13YJCZH065General Program of Science and Technology Development Project of Beijing Municipal Education Commission of China under Grant No.km201410005012Open Research Fund of Beijing Key Laboratory of Trusted Computing,Open Research Fund of Key Laboratory of Trustworthy Distributed Computing and Service(BUPT),Ministry of Education
文摘Online social networks have gradually permeated into every aspect of people's life.As a research hotspot in social network, user influence is of theoretical and practical significant for information transmission, optimization and integration. A prominent application is a viral marketing campaign which aims to use a small number of targeted infl uence users to initiate cascades of infl uence that create a global increase in product adoption. In this paper, we analyze mainly evaluation methods of user infl uence based on IDM evaluation model, Page Rank evaluation model, use behavior model and some other popular influence evaluation models in currently social network. And then, we extract the core idea of these models to build our influence evaluation model from two aspects, relationship and activity. Finally, the proposed approach was validated on real world datasets,and the result of experiments shows that our method is both effective and stable.
