Herd Clustering: A synergistic data clustering approach using collective intelligence

Traditional data mining methods emphasize on analytical abilities to decipher data, assuming that data are static during a mining process. We challenge this assumption, arguing that we can improve the analysis by vitalizing data. In this paper, this principle is used to develop a new clustering algorithm.Inspired by herd behavior, the clustering method is a synergistic approach using collective intelligence called Herd Clustering (HC). The novel part is laid in its first stage where data instances are represented by moving particles. Particles attract each other locally and form clusters by themselves as shown in the case studies reported. To demonstrate its effectiveness, the performance of HC is compared to other state-of-the art clustering methods on more than thirty datasets using four performance metrics. An application for DNA motif discovery is also conducted. The results support the effectiveness of HC and thus the underlying philosophy.     

具备反向学习和局部学习能力的磷虾群算法

针对磷虾群算法易陷入局部最优、收敛速度慢等缺点,提出了具备反向学习和局部学习能力的磷虾群算法。利用混沌映射和反向学习的思想初始化种群,根据算法迭代次数自适应调整学习维度,对精英个体进行反向学习,能有效保持种群的多样性,选取精英群体,通过自适应的Lévy飞行分布和改进的差分变异算子,提高种群的局部学习能力。这种新颖的元启发方式能加速收敛速度的同时可以保证磷虾群算法的鲁棒性。通过对8个基准函数进行仿真测试,实验结果表明：与最近的KH优化算法相比,该算法在收敛速度、收敛精度等方面得到明显改进。     

在校园网部署上网行为管理系统初探

本文从校园网安全存在的问题出发,分析部署上网行为管理系统的必要性和目前遇到的阻力,并给出解决方案。     

Herd behavior in social media: The role of Facebook likes,strength of ties,and expertise

When do social media users click on sponsored content or intend to visit the website at a later time? A qualitative comparative analysis (QCA) using arguments based on herd theory, strength of ties, and social distance shows that only “likes” from socially close and knowledgeable users can consistently generate click-through or view-through intentions. Considering social tie strength in a herd behavior context, the analysis of sufficient configurations for click- and view-through intentions provides a nuanced perspective on social media user behavior and social influence. For instance, click-through intention requires observing a “like” from a close person, while view-through intentions can also develop after observing “likes” from less close acquaintances, yet in the last case only if the user assumes the acquaintance is better informed regarding the sponsored content. In addition, a “like” from a close friend deemed better informed can even make a user click on a sponsored content that was not considered valuable before.     

手机产品的差异性对手机在线阅读行为的影响

本文研究的是手机产品的差异性对手机在线阅读行为的影响.通过分析用户的手机在线阅读行为,从“手机观感”,“手机性能”,“网络情况”三大基本影响因素进行探讨分析,通过相关理论研究以及问卷调查的方式得出手机性能、显示舒适度对手机在线阅读行为的影响最大,以及阅读平台的性能、版面设计、附加功能丰富度对其关系都具有一定的调节作用,且调节作用的大小不一.     

Potential and promise of online volunteering

The Internet is often perceived as yet another technological innovation that causes a further widening of the gap between rich and poor. However, one of the most interesting phenomena to come out of the Internet revolution is its utilization as a channel for social development. Volunteers working in task forces who were recruited through the net, work to improve the lives of many millions of people in need throughout the world. Some of these volunteers are ‘field workers’, working in physical proximity to those they are trying to help, while others may be sitting at home, using their computers to help needy populations many thousands of miles away. This paper studies this trend and advocates a model to explain the potential and promise of online volunteerism from the perspective of the volunteer. It is suggested that understanding the characteristics behind Internet volunteering from the perspective of the volunteer may enhance the positive potential of the Internet.     

Sex differences in purchasing textbooks online

In a series of studies, female students at an urban university tended to demonstrate fewer computer/Internet skills than male students, and their level of skill was a more consistent predictor of purchasing textbooks online. These results were not replicated at a rural state college with less computer-skilled students.     

Financial risk and its impact on new purchasing behavior in the online retail setting

This paper examines the effect of financial risk on perceptions of service quality and relationship-marketing quality in the online retail environment. Perceptions of financial risk were found to be negatively associated with service quality. In particular, a well-designed and attractive Web site was found to mitigate perceptions of financial risk during early trial-buy purchasing. Relationship-marketing quality was not affected by financial risk. This study adds to an enhanced understanding of how risk perceptions influence assessment of service quality and relationship-marketing quality. While the drivers of service quality and relationship-marketing quality have been examined extensively in the online setting, a surprising lack of research investigates the role of risk perceptions in the early stage, buy-trial purchasing behavior. This emerging area of research interest is deserving of more attention. Our findings provide valuable normative guidance to researchers interested in the affects of perceived risk (particularly financial risk) on new online shoppers, emphasizing the interdependency between Web site design and risk perceptions.     

Herd behavior in information security decision-making

Current information security behavior research assumes that lone individuals make a rational, informed decision about security technologies based on careful consideration of personally available information. We challenge this assumption by examining how the herd behavior influences users’ security decisions when coping with security threats. The results show that uncertainty about a security technology leads users to discount their own information and imitate others. We found that imitation tendency has a more substantial effect on security decisions than the personal perceived efficacy of the security technology. It is essential for researchers and managers to consider how the herd behavior effect influences users' security decisions.     

The mediating effect of website quality on Internet searching behavior

The study investigates the mediating effect of website quality on Internet searching behavior. The research model incorporates three dimensions, namely (1) Internet attitudes of the Internet users, (2) website quality, including the system quality, service quality and information quality, and (3) Internet searching behavior demonstrated by the general population of Internet users. A survey instrument was used to gather data to exam the relationships in the proposed model. The collected online surveys (n = 856) are used to test the relationships among the three dimensions expressed in the proposed structural equation model. The results show that website quality has a mediating effect on Internet searching behavior. It is also discovered that there is significant correlation between Internet attitudes and website quality. This study contributes empirical data to the predominantly theoretical literature on Internet searching behavior in general and Internet attitudes and website quality in particular. It is, to a certain extent, common sense that website quality has a positive impact on Internet searching behavior. This paper takes an important step forward by detailing how website quality affects the Internet searching behavior via enhancing the effect of users’ Internet attitudes.     

基于购买行为的随机生命周期易逝品库存策略

研究需求和商品生命周期均为随机的零售商库存管理问题,提出按照购买行为特征对需求进行分类,在不同类型的客户之间进行库存分配并允许缺货的库存策略.通过构建动态规划模型,求解出零售商最优库存策略,包括补货策略和库存分配策略.与先到先服务策略相比,该策略能显著提升零售商的利润,减少商品损坏的损失.     

基于改进磷虾群优化的中心极大化KFCM算法在IDS的应用*

由于核模糊C-均值算法(Kernel Fuzzy C-means,KFCM)随机选择初始聚类中心,易导致算法陷入局部最优,且算法在聚类中心较近或重合时,易产生一致性聚类结果。为解决以上问题,提出一种改进算法。改进算法对原目标函数进行重新定义,通过在目标函数中增加一项聚类中心约束项来调控簇间分离度,从而避免算法出现一致性聚类结果。利用改进磷虾群算法对基于新目标函数的KFCM算法进行优化,使算法不再依赖初始聚类中心,提高算法的稳定性。基于距离最大最小原则产生多组较优的聚类中心作为初始磷虾群体并在算法迭代过程中融合一种新的精英保留策略,从而确保算法收敛到全局极值。通过对个体随机扩散活动进行分段式logistic混沌搜索,提高算法全局寻优能力。使用KDD CUP 99入侵检测数据进行仿真实验表明,改进算法具有更好的检测性能,解决了传统的聚类算法在入侵检测中稳定性差,检测准确率低的问题。     

在线用户打分行为长记忆效应与信任关系研究

对在线打分行为的动态研究能够帮助深入理解社交网络用户集群行为和信任关系的演化机制,当前许多在线系统用户能够通过对物品进行打分传达自己的观点。通过去趋势波动分析法研究了用户打分行为在信任关系建立前后的长记忆效应,并通过随机化打分时间和信任时间建立零模型,最后进行用户打分行为异质性分析。采用Epinions数据集进行实证研究,结果表明用户打分的长记忆效应在信任关系建立前出现下降趋势（8.06%）,并于之后逐步回升（8.43%）,而在两个零模型中赫斯特指数分别稳定在0.5和0.6左右,且用户长记忆效应变动与用户度呈正相关,Pearson相关系数分别为0.9358和0.9278。该工作有助于深入理解用户集群行为和信任关系的动态演化机制。     

Teenagers in social virtual worlds: Continuous use and purchasing behavior in Habbo Hotel

Social virtual worlds (SVWs) have become important environments for social interaction. At the same time, the supply and demand of virtual goods and services is rapidly increasing. For SVWs to be economically sustainable, retaining existing users and turning them into consumers are paramount challenges. This requires an understanding of the underlying reasons why users continuously engage in SVWs and purchase virtual items. This study builds upon Technology Acceptance Model, motivational model and theory of network externalities to examine continuous usage and purchase intention and it empirically tests the model with data collected from 2481 Habbo users. The results reveal a strong relationship between continuous usage and purchasing. Further, the results demonstrate the importance of the presence of other users in predicting the purchase behavior in the SVW. Continuous SVW usage in turn is predicted directly by perceived enjoyment and usefulness while the effect of attitude is marginal. Finally, perceived network externalities exert a significant influence of perceived enjoyment and usefulness of the SVW but do not have a direct effect on the continuous usage.     

Exploring the impact of initial herd on overfunding in equity crowdfunding

Outcome of equity crowdfunding campaigns often exceeds original fundraising goals, leading to market inefficiency. This undesirable phenomenon of overfunding garnered little attention in past studies. Synthesizing extant literature on crowdfunding and herding, we identified the initial herd made visible by funding progress indicator as the main cause of overfunding. The impact of the initial herd can be quantified by three dimensions: maturity, intensity, and persistency. We then advance and validate a research model for examining how dimensions of the initial herd affect overfunding in equity crowdfunding. Findings from this study can shed light on plausible remedies for the overfunding issue.     

Exploring the value of purchasing online game items

The aims of this paper are to: (1) modify the theory of consumption values in order to investigate online game users’ perceived value of purchasable game items, and (2) develop a new construct – the “integrated value of purchasing game items” – based on the modified theory of consumption values. We found that the enjoyment, character competency, visual authority, and monetary values are appropriate for describing how online game users perceive the value of game items. Utilizing second order analysis, the “integrated value of purchasing game items” was developed. To show the validity of the new construct, we developed a research model and tested it using the results of 327 valid questionnaires. Results revealed that the new construct is statistically significant in affecting users’ intention to purchase game items.     

Communicative practices in an online financial forum during abnormal stock market behavior

We offered a new perspective on communication in computer-mediated environments by examining communicative practices used during discussion in an online financial forum about a listed company that was experiencing abnormal trading behavior. The analysis of communicative practices showed how the linguistic nature of the online forum allowed covert manipulation of inexperienced and unsuspecting forum participants - against all the norms and rules governing online financial forums. Based on our theoretical understanding of communicative practices, we found a discrepancy between the expectations of and the actual divergence from an ideal communicative environment free from distortion and manipulation that is normally assumed to govern online forums. The discrepancy, which remains largely concealed, can be seen as a major factor in the negative implications of online financial forums.     

基于贝叶斯分类的网上书店潜在用户挖掘

以网上书店为例,利用贝叶斯分类预测技术,进行了发现潜在客户群体的研究,用随机选取的10组样本进行试验预测,预测准确率达96.5%,表明了该算法是有效的。     

Netiquette within married couples: Agreement about acceptable online behavior and surveillance between partners

The internet has become an integral part of many people’s everyday lives. It is unclear what its role is in maintaining intimate offline relationships and whether the use of the internet might cause conflicts between partners about what constitutes acceptable online behavior. An online survey of 920 married couples in the UK who used the internet investigated whether partners have similar netiquettes. There were high levels of agreement between married partners about the unacceptability of online infidelities; similarly they agreed more than two random individuals about the acceptability of entertainment activities which, in excess, might be addictive. Partners further showed high correspondence in surveillance behavior. Women were more concerned about their own and their partner’s behavior and were more likely to monitor their partner’s online activities. These findings suggest that a netiquette is developed and consciously or subconsciously negotiated within intimate relationships. Nevertheless, traditional gender differences as regards risk perception still hold; women are more likely to problematies their own and their partners behaviors.     

The role of atmospheric cues in online impulse-buying behavior

This study extends a stimulus–organism–response (S–O–R) model to include impulse-buying behavior, which plays a vital role in electronic shopping but has not gained much attention in e-commerce research. Grounding our research in environmental psychology, we test the effects of virtual atmospheric cues on online impulse-buying behavior and spending, via a consumer survey. The study applies elaborated mediating variables (shopping enjoyment and impulsiveness) to develop a structural model linking three categories of atmospheric cues of an electronic store (content, design, and navigation) to approach behavior variables (impulse-buying behavior and expenditure). The results support the validity of the S–O–R model in the context of online impulse-buying behavior and show a significant positive effect of two dimensions of virtual atmospheric cues (design and navigation).