基于用户兴趣和推荐信任域的微博推荐

向用户推荐其感兴趣的微博,是改善用户体验的重要途径.为使推荐的微博更加符合用户的兴趣和品味,提出的微博推荐方法不仅考虑用户自身的特点,而且还考虑所在社区对微博的评价.在技术实现上,采用支持向量机进行文本分类,以便发现用户的兴趣偏好;通过多维Newman算法进行用户社区的发现,并将社区视为推荐信任域.最后采用改进的协同过滤算法综合用户兴趣偏好和推荐信任域进行微博推荐,以此提高微博推荐的质量.实验结果表明,提出的微博推荐方法是切实有效的.     

融合标签关联关系与用户社交关系的微博推荐方法

通过分析微博特点及现有微博推荐算法的缺陷,提出一种融合了标签间关联关系与用户间社交关系的微博推荐方法.采用标签检索策略对未加标签和标签较少的用户进行加标,构建用户-标签矩阵,得到用户标签权重,为了解决该矩阵中稀疏的问题,通过挖掘标签间的关联关系,继而更新用户-标签矩阵.考虑到多用户之间社交关系对挖掘用户兴趣并进行微博推荐的重要性,构建用户-用户社交关系相似度矩阵,并与更新后的用户-标签矩阵进行迭代,得到最终的用户兴趣并进行相关推荐.实验证明了该算法针对微博信息推荐是有效的.     

基于特征分析的微博用户兴趣发现算法

本文在综合兴趣模型研究现状的基础上,结合微博数据集对微博用户的特征进行分析,建立微博用户兴趣模型,并提出基于微博用户兴趣模型的发现算法。实验结果表明,本文提出的算法能很好的发现微博用户的兴趣,提高推荐系统的质量。     

基于核心标签的可重叠微博网络社区划分方法

针对传统微博社区发现算法内聚低重叠度不可控制等问题,以自顶向下的策略,提出一种基于核心标签的可重叠微博社区发现策略Tag Cut.先利用用户标签的共现关系及逆用户频率对标签进行加权,并基于标签之间的内联及外联关系并将用户的标签进行扩充,然后在整体社区中提取包含某一标签的用户作为临时分组并利用评价函数评估划分的优劣,最后选出最合适的核心标签根据其对应分组与其他分组距离的远近来决定将其划分为新的分组还是并入其他分组.用此策略反复迭代直到满足要求.该算法划分的组由若干个拥有核心标签的分组组成且综合利用微博用户已声明的及隐含的兴趣、用户之间的关注规律、结果的实用性对划分结果进行修正.经真实数据实验表明该方法内聚高社区重叠度可控且拥有实际意义.     

社会化标签系统中基于本体的个性化推荐方法研究

【目的】捕捉用户兴趣的动态性变化,优化个性化信息推荐效果。【应用背景】高效的个性化信息推荐方法可以根据用户行为特征主动为用户提供合适的信息资源,使信息的获取和利用更加快捷、准确。【方法】以"新浪微博"为例,通过挖掘用户及其关注者的微博数据,提取标签,计算二者兴趣相似度及亲密度,确定用户兴趣标签并优化标签描述,从而构建用户个性化"轻量级"本体,使得语义网资源能够准确地投放到用户界面。【结果】有效缓解了信息爆炸式增长所造成的"信息迷航"现象。【局限】微博数据中的杂音(广告转发、多语言描述)、数据不充分等,可能影响标签提取的准确性。     

基于本体论的个性化影片推荐技术研究

个性化影片推荐服务是解决目前网络及家庭数字电视应用中影片资源迅速增长,用户"信息迷航"的有效方法.针对影片点播应用,给出了个性化影片推荐服务的体系结构、影片数据建模、用户兴趣偏好模型进行了研究,实现无需用户输入传统推荐方法所需相关个性兴趣信息即可返回与用户当前兴趣相关的影片推荐列表,提出了基于本体论的影片模型,并建立用户兴趣偏好模型,给出了对推荐过程中结合用户信息反馈对推荐结果进行自适应的调整算法.     

基于图神经网络的视频推荐系统

提出了一种基于图神经网络的视频推荐模型,将用户的视频观看序列型行为建模为图结构,用结点代表用户与视频,用边代表行为,引入两种类型的向量传播方法分别对用户的长期兴趣与短时兴趣进行建模。其中,通过用户结点与视频结点的双向传播刻画长期兴趣,借助视频结点切换关系的单向传播刻画短时兴趣,并通过多层向量传播实现对图上高阶邻接信息的捕捉。在一个真实世界的视频网站观看数据集上的实验表明,提出的方法与现有最佳方法相比,其推荐精准度得到了有效提升。进一步的实验表明,该方法能够有效缓解数据稀疏性的问题。     

一种基于用户隐式反馈的微博信息过滤方法

微博平台的兴起革新了人们的互动方式,给人们获取信息带来了极大便利.然而,在信息超载的环境下,人们需要花费大量的时间从许多冗余的微博信息中寻找自己感兴趣的信息,剔除无用信息.针对该问题,本文设计了一种新的方法对用户的微博信息进行过滤.该方法在传统方法基础上增加用户反馈环节;同时,考虑用户兴趣随时间变化的特点,在进行信息过滤时考虑时间对兴趣度的影响.该方法为微博信息个性化过滤提供了一种新思路.     

基于行为延迟共享网络的个性化商品推荐方法

针对电商平台难以利用历史浏览行为进行个性化商品推荐的问题,该文提出了一种行为延迟共享网络模型(BDSN),充分结合历史浏览信息,对用户进行精准浏览推荐.该模型提出行为延迟门控循环神经单元(BD-GRU),将历史浏览时间间隔作为用户活跃度因子,对神经元状态进行更新,用于计算用户的兴趣表示.为了提高向量表示的一致性,该模型提出共享参数网络,将用户侧和商品侧的表示向量收敛到统一空间,解决个性化商品推荐点击率预估问题.并在真实数据集上进行实验,结果表明,BDSN模型在验证集上的AUC指标和损失函数均处于最优,在测试集上的AUC指标相较基本模型提高37％,能够有效提升商品推荐的准确性.     

总论

0513052数字图书馆个性化服务用户模型研究[刊,中]/宋丽哲//北京理工大学学报.—2005,25(1).—58-62(L2)提出了一种数字图书馆个性化服务用户模型构架,并对实现过程中的几个关键问题,包括用户模型表示方法,用户模型的建立以及更新算法进行了详细论述。提出了基于本体论的空间向量用户模型表示方法,建立了简单的数字图书馆领域本体,以空间向量表示用户模型,以本体概念作为向量的特征项;采用支持向量机分类算法和无监督聚类算法相结合提取用户兴趣;在用户模型更新方法上,采用渐进遗忘和滑动窗口相结合的方法实现用户兴趣概念的漂移。参7     

Social network users clustering based on multivariate time series of emotional behavior

It is known that the social network is an excellent source for gathering the emotions of people. There are thousands of micro-blogs posted in every second and every micro-blog that may contain a variety of user＇s emotions. The users＇ collective emotional behaviors are with great impacts on today＇s societies, so it is good to find groups for society management based on users＇ emotional behavior. This article focuses on analyzing multivariate emotional behavior of users in social network and the goal is to cluster the users from a fully new perspective-emotions. The following tasks are completed： firstly, the multivariate emotion of Chinese micro-blog with vector is analyzed, and multivariate time series to describe the user＇s emotional behavior are constructed. Seconedly, considering principal component analysis （PCA） in similarity and distance similarity, the similarity of the multivariate emotion time series is measured. The contribution could be summarized as follows： groups of users though different emotions in social network are discovered. The emotional fluctuation and intensity of users are considered as well. Experiment in clustering effectively illustrates the emotional behavior characteristics of the Users in different groups.     

基于用户主题模型的微博用户兴趣挖掘(英文)

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.     

基于上下文相似度和社会网络的移动服务推荐方法

针对传统的基于协同过滤的移动服务推荐方法存在的数据稀疏性和用户冷启动问题,提出一种基于上下文相似度和社会网络的移动服务推荐方法(Context-similarity and Social-network based Mobile Service Recommendation,CSMSR).该方法将基于用户的上下文相似度引入个性化服务推荐过程,并挖掘由移动用户虚拟交互构成的社会关系网络,按照信任度选取信任用户;然后结合基于用户评分相似度计算发现的近邻,分别从相似用户和信任用户中选择相应的邻居用户,对目标用户进行偏好预测和推荐.实验表明,与已有的服务推荐方法TNCF、SRMTC及CF-DNC相比,CSMSR方法有效地缓解数据稀疏性并提高推荐准确率,有利于发现用户感兴趣的服务,提升用户个性化服务体验.     

Wayfinding and art viewing by users of a mobile system and a guidebook

This study examined location-based service users' perception and behaviour in a large, multi-story commercial complex, by comparing wayfinding, acquired spatial knowledge and art evaluation by people who viewed artworks using a mobile art-tour system and by people who used a paper-based guidebook. Results showed that mobile users made wayfinding errors because of difficulties in understanding navigational directions offered by the system, whereas guidebook users made errors because of difficulties in locating their positions on maps, especially when they moved between floors. Mobile users' sketch maps similarly showed errors of locating artworks on incorrect floors. In terms of scene recognition, mobile and guidebook users performed comparably, indicting that they acquired equivalent levels of landmark knowledge. But in terms of memory retention, mobile users tended to forget about visited artworks as time passed. Mobile and guidebook users did not differ in the evaluation of artworks, but differed in the viewing of offered contents. Mobile users, especially those with low interest in art, tended not to view contents offered in the formats of photographs and movies, whereas guidebook users viewed all contents in the guidebook. Possible effects of the degree of active involvement in wayfinding and content viewing, and issues about raising motivation of low-interest people so that they interact with the terminal device, and preferably gain interest in guided activities, are discussed.     

Privacy measurement method using a graph structure on online social networks

Recently, with an increase in Internet usage, users of online social networks (OSNs) have increased. Consequently, privacy leakage has become more serious. However, few studies have investigated the difference between privacy and actual behaviors. In particular, users' desire to change their privacy status is not supported by their privacy literacy. Presenting an accurate measurement of users' privacy status can cultivate the privacy literacy of users. However, the highly interactive nature of interpersonal communication on OSNs has promoted privacy to be viewed as a communal issue. As a large number of redundant users on social networks are unrelated to the user's privacy, existing algorithms are no longer applicable. To solve this problem, we propose a structural similarity measurement method suitable for the characteristics of social networks. The proposed method excludes redundant users and combines the attribute information to measure the privacy status of users. Using this approach, users can intuitively recognize their privacy status on OSNs. Experiments using real data show that our method can effectively and accurately help users improve their privacy disclosures.     

A time-aware query-focused summarization of an evolving microblogging stream via sentence extraction

With the number of social media users ramping up, microblogs are generated and shared at record levels. The high momentum and large volumes of short texts bring redundancies and noises, in which the users and analysts often find it problematic to elicit useful information of interest. In this paper, we study a query-focused summarization as a solution to address this issue and propose a novel summarization framework to generate personalized online summaries and historical summaries of arbitrary time durations. Our framework can deal with dynamic, perpetual, and large-scale microblogging streams. Specifically, we propose an online microblogging stream clustering algorithm to cluster microblogs and maintain distilled statistics called Microblog Cluster Vectors (MCV). Then we develop a ranking method to extract the most representative sentences relative to the query from the MCVs and generate a query-focused summary of arbitrary time durations. Our experiments on large-scale real microblogs demonstrate the efficiency and effectiveness of our approach.     

Mining microblog user interests based on TextRank with TF-IDF factor

It is of great value and significance to model the interests of microblog user in terms of business and sociology. This paper presents a framework for mining and analyzing personal interests from microblog text with a new algorithm which integrates term frequency-inverse document frequency (TF-IDF) with TextRank. Firstly, we build a three-tier category system of user interest based on Wikipedia. In order to obtain the keywords of interest, we preprocess the posts, comments and reposts in different categories to select the keywords which appear both in the category system and microblogs. We then assign weight to each category and calculate the weight of keyword to get TF-IDF factors. Finally we score the ranking of each keyword by the TextRank algorithm with TF-IDF factors. Experiments on real Sina microblog data demonstrate that the precision of our approach significantly outperforms other existing methods.     

A QoE centric distributed caching approach for vehicular video streaming in cellular networks

Distributed caching‐empowered wireless networks can greatly improve the efficiency of data storage and transmission and thereby the users' quality of experience (QoE). However, how this technology can alleviate the network access pressure while ensuring the consistency of content delivery is still an open question, especially in the case where the users are in fast motion. Therefore, in this paper, we investigate the caching issue emerging from a forthcoming scenario where vehicular video streaming is performed under cellular networks. Specifically, a QoE centric distributed caching approach is proposed to fulfill as many users' requests as possible, considering the limited caching space of base stations and basic user experience guarantee. Firstly, a QoE evaluation model is established using verified empirical data. Also, the mathematic relationship between the streaming bit rate and actual storage space is developed. Then, the distributed caching management for vehicular video streaming is formulated as a constrained optimization problem and solved with the generalized–reduced gradient method. Simulation results indicate that our approach can improve the users' satisfaction ratio by up to 40%. Copyright © 2015 John Wiley & Sons, Ltd.     

移动互联网用户隐私信息检测保护技术研究及应用

智能终端及应用作为“大连接”中的重要节点和业务载体，直接或间接接触大量用户敏感隐私信息，近年来，App强制授权、过度索权、超范围收集个人信息的现象大量存在， 违法违规使用个人信息的问题十分突出，用户隐私泄露的情况愈演愈烈，安全及隐私问题引发社会广泛关注。本论文根据不同源头的APP隐私安全风险全面梳理排查，创新提出“静态权限检测+动态行为特征+网络DPI智能分析”的隐私信息检测防护技术体系，实现了敏感权限智能分析、违规索权动态监控、隐私泄露探测预警、敏感信息深度追踪，确保移动应用APP安全、可信、可控，保障了业务单位和用户隐私安全权益。