微博网络中用户特征量和增长率分布的研究

根据新浪微博的实际用户数据,发现粉丝数、关注数和微博数3个特征量的分布,都存在双段幂律分布现象,不同类型用户特征量的分布具有差异性.使用双帕累托对数正态(DPLN)分布对数据进行拟合,相比对数正态分布和幂律分布,可以得到更优的效果.用户活跃时间服从指数分布,不同活跃时间的3个用户特征量都近似服从对数正态分布;用户特征量的增长率服从对数正态分布,且与特征量自身的规模无关,这些特征与双帕累托对数正态分布模型一致,从而使用这一模型可以很好地解释粉丝数、关注数和微博数分布特性的形成机制.     

面向领域微博权威性人物分析技术与研究

为了精准地获取微博领域权威性用户排序列表,在现有研究成果的基础上以新浪微博为研究对象,手动构建领域词典,面向微博领域提出了领域微博权威性用户度量模型(domains-microblog authoritative user ranking model,DMAURank).该模型使用基于PageRank与用户行为权值相结合的评价方法.将实验结果与用户平均转发数排名和粉丝数排名进行对比,发现基于PageRank与用户行为权值相结合的用户被关注度有效降低了粉丝数目的过大影响.     

新浪微博用户及其微博特征分析

基于新浪微博用户之间的关注关系网络,分析了衡量微博用户影响力的三个指标--粉丝数、User PR值以及用户活跃度,发现粉丝数分布和User PR值分布均服从幂律分布,活跃度分布不同于前两种分布。分别对三种排名靠前的用户及其发布的微博进行分析,发现排名靠前的用户中,User PR值的认证用户多于粉丝数;活跃度排名靠前的用户在广告营销活动中受到广泛的青睐;新浪微博用户乐于转发和评论他人的微博,微博中嵌入了大量的图片、视频和链接。     

基于HRank的微博用户影响力评价

针对微博社交网络平台中的用户影响力评价问题,提出了一种基于HRank的评价算法。该算法将评价科学家科研绩效影响力的判定参数H指数引入进来,构造出能反映用户影响覆盖度的粉丝H指数和用户微博受追捧程度的微博被转发H指数,以分别表征用户的静态特征和在微博平台上的动态行为特征。在此基础上,结合粉丝H指数和微博被转发H指数构建出对用户影响力进行综合评价的HRank模型。粉丝数与用户影响力的相关性不是很强,同样数据集下相对PageRank,HRank用户影响力模型与新浪用户影响力官方排名更为接近,可有效实现对微博用户影响力的客观评判。     

面向微博影响力的社交网络特征分析

社交网络的影响力与其自身的结构特征密切相关。基于新浪微博的数据,对用户的粉丝数、关注数的分布及这些特征之间的关系进行分析,发现用户的粉丝数、关注数、微博数都符合幂律分布;探讨了节点之间的距离特征,发现并证明了微博网络中存在着“小世界”现象;研究了节点之间的链接形成问题,发现链接的形成满足三元闭包原理。以上三方面研究结果,对于探索微博影响力同底层社交网络结构特征的关系、设计微博影响力控制机制具有重要的意义。     

微博数据通用抓取算法

目前常用的网络爬虫和基于微博API抓取数据的算法很难满足舆情系统对微博数据的需求。为此,提出一种模拟浏览器登录微博抓取网页数据的算法,以方便地获取任意微博用户网页上的所有数据。通过微博用户之间的关系构建用户网络,并通过该网络发现新用户。为获取微博上有质量的数据,建立一个完整的数学模型,根据用户的发帖数、发帖频率、粉丝数、转发数、评论数等因素来计算用户影响力,以影响力为主要因子构建优先队列,使得影响力越大的用户数据采集频率越高,同时计算时间间隔以兼顾非活跃用户的数据获取。实验结果表明,该算法具有通用性强、完全无需人工干预、获取信息的质量高、速度快等优点。     

一种面向微博用户的标签推荐方法

首先对国内微博平台的信息进行了综合分析,主要介绍了微博信息的定义,在错综复杂的微博信息中哪些信息比较重要,以及这些微博信息包含哪些详细的内容,是如何组织的。然后选取新浪微博平台作为研究对象,利用新浪微博API设计了爬虫程序,抽取用户信息;以用户的关注人数、粉丝数和发布的微博数为标准对用户信息进行了定量分析。最后根据分析结果,针对不同特征的用户群体提出了相应的标签推荐方法。     

一种面向微博用户的标签推荐方法

首先对国内微博平台的信息进行了综合分析,主要介绍了微博信息的定义,在错综复杂的微博信息中哪些信息比较重要,以及这些微博信息包含哪些详细的内容,是如何组织的。然后选取新浪微博平台作为研究对象,利用新浪微博API设计了爬虫程序,抽取用户信息;以用户的关注人数、粉丝数和发布的微博数为标准对用户信息进行了定量分析。最后根据分析结果,针对不同特征的用户群体提出了相应的标签推荐方法。     

微博用户的时间行为研究

人类日常行为活动在生活的各个方面普遍存在。个体行为活动类型多样且个体之间行为差异明显,人类行为显示出高度复杂性。利用微博用户发布微博所产生的时间数据来研究用户的时间行为模式,研究发现个体用户时间间隔的统计特征主要服从幂律、指数和双模3种分布函数;并提出了基于任务队列的个体用户行为动力学模型,解释了用户发布微博的时间间隔分布特征。     

基于相似度的微博社交网络的社区发现方法

作为一种新兴的社交媒体,微博由于其信息的简短性、实时性和公开性,在短短4年内已积累数以亿计的用户并且数量还在迅速增长,由此带来的社会影响日益广泛.对微博用户关系网络进行社区发现具有重要的理论和实际意义.根据微博网络的有向性及建立关注关系的随意性等特点,提出一种基于共同关注和共同粉丝的微博用户相似度,定义此相似度的模块化函数,依据贪心算法思想设计出基于此模块化函数最大化的社区发现方法,并在此基础上将该方法推广到具有标签信息的微博网络中.应用该方法处理了3个真实的微博用户关系网络数据,结果表明该方法可以有效地发掘微博用户关系网络中的社区结构.     

一种基于OUS的最大频繁项集挖掘算法

分析实际应用中有效访问序列的特点,提出了一种采用自底向上策略快速挖掘最大频繁项集的OUS算法。该算法首先对用户项集进行重叠操作统计浏览次数,然后合并,依据用户给出的最小支持度删除原项集中的非频繁页面元素,并对两两用户项集筛选生成候选频繁项集,最后扫描数据库,统计各个候选频繁项集的支持度计数。实验结果表明,该算法能有效地发现用户最大频繁项集。     

Spatiotemporal feature mining algorithm based on multiple minimum supports of pattern growth in Internet of Things

The temporal and spatial characteristics of users are involved in most Internet of Things (IoT) applications. The spatial and temporal movement patterns of users are the most direct manifestation of the temporal and spatial characteristics. The user’s interests, activities, experience and other characteristics are reflected by mobile mode. In view of the low clustering efficiency of moving objects in convergent pattern mining in the IoT, a spatiotemporal feature mining algorithm based on multiple minimum supports of pattern growth is proposed. Based on the temporal characteristics of user trajectories, frequent and asynchronous periodic spatiotemporal movement patterns are mined. Firstly, the location sequence is modeled, and the time information is added to the model. Then, a mining algorithm of asynchronous periodic sequential pattern is adopted. The algorithm is based on multiple minimum supports of pattern growth. According to multiple minimum supports, the sequential pattern of asynchronous period is mined deeply and recursively. Finally, the proposed method is validated and evaluated by Gowalla dataset, in which the user characteristics are truly reflected. It is shown by the experimental results that the average pointwise mutual information (PWI) of the proposed algorithm reaches 0.93. And the algorithm is proved to be effective and accurate.

     

Social models in open learning object repositories: A simulation approach for sustainable collections

Learning object repositories (LOR) are digital collections of educational resources and/or metadata aimed at facilitating reuse of materials worldwide. In open repositories, resources are made available at no cost, representing a case of information sharing with an implicit and diffuse social context. In such settings, quality control is in many cases based in some form of community filtering that provides a reliable basis for ranking resources when repositories reach a critical mass of users. However, there have been numerous repository initiatives and projects and many of them did not reached a significant degree of actual usage and growth that made them sustainable in the long term. In consequence, finding models for sustainable collections is a key issue in repository research, and the main problem behind that is understanding the evolution of successful repositories. This in turn requires analyzing experimental models of the behavior of their users that are coherent with the available evidence on their structure and growth patters. This paper provides a partial model for such behavior based on existing reported evidence and on the examination of patterns in a large and mature repository. Agent-based simulation was chosen to allow for contrasting configurations with different parameters. Simulations were devised with the RePast framework and the resulting model implementation constitutes an initial baseline for future studies aimed at contrasting empirical data on repository usage with their community setting. The model described accounts for known user contribution patterns and it is coherent with the implicit social network structure found in an existing large LOR.     

MIMO系统中基于非正交多址接入的功率分配算法研究

随着移动通信技术的发展,通信服务已变成人类日常生活中不可或缺的部分.尤其是近年来各类智能终端的大众化,使得接入无线通信的用户数和人们对通信服务的需求均呈爆炸式的增长.但现如今可用的频谱资源是有限的,且传统的正交多址接入系统的用户接入数受限,很难满足用户日益增长的需求.非正交多址接入允许在同一时频资源上复用多个用户,极大...     

The dynamic competitive recommendation algorithm in social network services

As the number of Twitter users exceeds 175 million and the scale of social network increases, it is facing with a challenge to how to help people find right people and information conveniently. For this purpose, current social network services are adopting personalized recommender systems. Existing recommendation algorithms largely depend on one of content-based algorithm, collaborative filtering, or influential ranking analysis. However, these algorithms tend to suffer from the performance fluctuation phenomenon in common whenever an active user changes, and it is due to the diversities of personal characteristics such as the local social graph size, the number of followers, or sparsity of profile content. To overcome this limitation and to provide consistent and stable recommendation in social networks, this study proposes the dynamic competitive recommendation algorithm based on the competition of multiple component algorithms. This study shows that it outperforms previous approaches through performance evaluation on actual Twitter dataset.     

基于模式匹配度的用户移动规则挖掘及位置预测方法研究

传统的移动用户位置预测方法由于模式支持度计算方式不合理,存在预测精度偏低的问题。为此,提出了一种基于模式匹配度的用户移动规则挖掘及位置预测方法,并将其用于移动通信系统中,以基台覆盖范围网格为单元的用户位置预测。具体包括三个步骤：通过图的遍历挖掘用户移动模式、基于用户移动模式生成用户移动规则和依据用户移动规则进行位置预测。实验分析使用10个批次轨迹数据进行用户移动规则挖掘,结果表明,该方法挖掘出的用户移动规则数少、支持度高和置信度高,具有高精度的优点。     

Are you connected? Evaluating information cascades in online discussion about the #RaceTogether campaign

In the context of online discussion about the recent Starbucks' “Race Together” cup campaign, this study aims to explore the central users in the online discussion network on Twitter and the factors contributing to a user's central status in the network. A social network analysis of 18,000 unique tweets comprising 26,539 edges and 14,343 Twitter users indicated five types of central users: conversation starter, influencer, active engager, network builder, and information bridge. Moreover, path analysis revealed that the number of people a Twitter user follows, the number of followers a user has, and the number of tweets a user generates within a time period helped a user increase his/her “indegree” connections in the network, which, together with one's “out-degree” connections in the network, propelled a user to become a central figure in the network.     

社交网络中基于K核分解的意见领袖识别算法

针对在社交网络中挖掘意见领袖时存在的计算复杂度高的难题,提出了一种基于K核分解的意见领袖识别算法CR.首先,基于K核分解方法获取社交网络中的意见领袖候选集,以缩小识别意见领袖的数据规模;然后,提出包括位置相似性和邻居相似性的用户相似性的概念,利用K核值、入度数、平均K核变化率和用户追随者个数计算用户相似性,并根据用户相...     

基于序列模式的用户浏览行为提取与分析

当今互联网所提供的功能和服务越来越多,Web内容也越来越丰富,移动应用越来越流行。然而,复杂的Web服务应用对用户提出了更高的要求,给用户浏览带来了很多问题,很多时候用户会感到无所适从。文中提出基于用户浏览序列模式的用户行为提取与分析方法。该方法可以分为浏览模式分析和用户聚类两部分。在浏览模式分析时,首先根据用户行为数据得到浏览序列,然后运用序列模式挖掘PrefixSpan算法获取用户习惯的浏览模式,最后把分析获取的用户浏览模式应用到Web浏览中,为不同的用户需求提供个性化的服务。在用户聚类时,运用层次聚类方法按照浏览模式的相似性对用户进行聚类,以分析用户的不同属性（如年龄、职业、学历等）对用户浏览模式的影响。实验结果表明,文中采用的PrefixSpan算法和层次聚类方法在用户浏览模式分析和研究方面具有很好的可行性和有效性。     

Understanding the popular users: Following,affiliation influence and leadership on GitHub

Context: the ability to follow other users and projects on GitHub has introduced a new layer of open source software development participants who observe but do not contribute to projects. It has not been fully explored how following others influences the actions of GitHub users. Objective: this paper studies the motivation behind following (or not following) others and the influence of popular users on their followers. Method: a mixed methods research approach was used including a survey of 800 GitHub users to uncover the reasons for following on GitHub and a complementary quantitative analysis of the activity of GitHub users to examine influence. Our quantitative analysis studied 199 popular (most followed) users and their followers. Results: we found that popular users do influence their followers by guiding them to new projects. As a user’s popularity increases, so does their rate of influence, yet the same is not true for a popular user’s rate of contribution. Conclusions: these results indicate that a new type of leadership is emerging through GitHub’s following feature and popularity can be more important than contribution in influencing others. We discuss implications of popularity and influence and their impact on social structure and leadership on OSS projects.