《Social Network Data Analytics》：社会学视角下的信息技术算法详解

需积分: 16 93 浏览量更新于2024-07-23 收藏 4.54MB PDF 举报

《<hibernate-mapping>：社会网络数据分析》这本书是一部独特而深入的数据分析著作，由CharuC.Aggarwal编辑，特别关注社会学视角在数据分析中的应用。它将理论与实践相结合，探讨了在社交网络数据领域中的复杂问题和算法。Aggarwal博士来自IBM Thomas J. Watson Research Center，位于美国纽约州霍索恩的19 Skyline Drive。本书的核心内容围绕Social Network Data Analytics展开，涵盖了从理论框架到实际案例的全面讲解，旨在帮助读者理解如何利用社会学原理和技术工具来解析和解读社交网络数据。作者提供了丰富的实证研究和案例分析，让读者能够掌握如何通过Hibernate等数据库映射语言进行数据管理和查询，这对于处理大规模、动态的社会网络数据至关重要。 ISBN 978-1-4419-8461-6和978-1-4419-8462-3分别标识了不同的版本，同时该书还被赋予了DOI 10.1007/978-1-4419-8462-3，这表明它是Springer New York出版的一部学术作品，并遵循了严格的版权规定。Library of Congress Control Number 2011922836进一步确认了本书在美国的版权登记。读者应注意，未经Springer Science+Business Media LLC的书面许可，不得全文翻译或复制，除非用于学术评论或学术分析。此外，书中提及的任何商标、服务标记或类似术语的使用，并不意味着作者对它们的专属性有任何评价，而是为了说明相关概念和行业的通用实践。《<hibernate-mapping>：社会网络数据分析》是一本适合数据分析师、社会科学家以及计算机科学专业人员的必读之作，它不仅提供了扎实的理论基础，还为在实际工作中处理社交网络数据提供了解决方案和实用工具。对于希望在社交网络数据分析领域深化理解并应用Hibernate技术的读者来说，这本书是提升技能和知识的重要资源。

2 SOCIAL NETWORK DATA ANALYTICS

Facebook and LinkedIn. Such social networks have lead to a tremendous ex-

plosion of network-centric data in a wide variety of scenarios. Social networks

can be de¿ned either in the context of systems such as Facebook which are ex-

plicitly designed for social interactions, or in terms of other sites such as Flickr

which are designed for a different service such as content sharing, but which

also allow an extensive level of social interaction.

In general, a social network is de¿ned as a network of interactions or re-

lationships, where the nodes consist of actors, and the edges consist of the

relationships or interactions between these actors. A generalization of the idea

of social networks is that of information networks, in which the nodes could

comprise either actors or entities, and the edges denote the relationships be-

tween them. Clearly, the concept of social networks is not restricted to the

speci¿c case of an internet-based social network such as Facebook; the prob-

lem of social networking has been studied often in the ¿eld of sociology in

terms of generic interactions between any group of actors. Such interactions

may be in any conventional or non-conventional form, whether they be face-to-

face interactions, telecommunication interactions, email interactions or postal

mail interactions.

The conventional studies on social network analysis have generally not fo-

cussed on online interactions, and have historically preceded the advent and

popularity of computers or the internet. A classic example of this is the study

of Milgram [18] in the sixties (well before the invention of the internet), who

hypothesized the likelihood that any pair of actors on the planet are separated

by at most six degrees of separation. While such hypotheses have largely re-

mained conjectures over the last few decades, the development of online social

networks has made it possible to test such hypotheses at least in an online

setting. This is also referred to as the small world phenomenon. This phe-

nomenon was tested in the context of MSN messenger data, and it was shown

in [16] that the average path length between two MSN messenger users is 6.6.

This can be considered a veri¿cation of the widely known rule of “six degrees

of separation” in (generic) social networks. Such examples are by no means

unique; a wide variety of online data is now available which has been used

to verify the truth of a host of other conjectures such

as that of shrinking

diameters[15] or preferential attachment. In general, the availability of mas-

sive amounts of data in an online setting has given a new impetus towar ds a

scienti¿c and statistically robust study of the ¿eld of social networks.

The shrinking diameter conjecture hypothesizes that the diameters of social networks shrink in spite of

the addition of new nodes, because of an increase in the density of the underlying edges. The preferential

attachment conjecture hypothesizes that new nodes and edges in the social networks are more likely to be

attached to the dense regions of the network.

An Introduction to Social Network Data Analytics 3

This data-centric impetus has lead to a signi¿cant amount of research, which

has been unique in its statistical and computational focus in analyzing large

amounts of online social network data. In many cases, the underlying insights

are applicable to the conventional social network setting as well. Before dis-

cussing the research topics in more detail, we will brieÀy enumerate the differ-

ent settings for social network analysis, and speci¿cally distinguish between

the conventional and non-conventional scenarios. Speci¿cally, these different

settings are as follows:

The most classical de¿nition of a social network is one which is based

purely on human interactions. This is the classical study of social net-

works in the ¿eld of sociology. These studies have traditionally been

conducted with painstaking and laborious methods for measuring in-

teractions between entities by collecting the actual data about human

interactions manually. An example is the six-degrees-of-separation ex-

periment by Milgram [18], who used postal mail between participants in

order to test whether two arbitrary actors could be connected by a chain

of 6 edges with the use of locally chosen forwards of the mail. Such

experiments are often hard to conduct in a completely satisfactory way,

because the actors in such experiments may have response rates which

cannot be cleanly modeled in terms of social interaction behavior. An

example is the case of the Milgram experiment, in which the results have

often been questioned [14] because of the low forward rate of the letters

which never reached the target. Furthermore, such social experiments

are often biased towards high status targets in order to ensure likelihood

of logical forwards. However, these results have eventually been ac-

cepted at least from a qualitative perspective, even though the rule of

six degrees may not be precisely correct, depending upon the nature of

the network which is being studied. Nevertheless, the “small world phe-

nomenon” de¿nitely seems to be correct, since the diameters of most

such networks are relatively small.

The social analysis of such networks has also been modeled in the ¿eld

of cognitive science, where the cognitive aspects of such interactions are

utilized for analytical purposes. Much of the research in the traditional

¿eld of social networks has been conducted from this perspective. A

number of books [7, 24, 25] provide an understanding of this perspec-

tive. However, this work does not discuss the data-centric issues which

are common to online and internet-enabled social networks.

A number of technological enablers such as telecommunications, elec-

tronic mail, and electronic chat messengers (such as Skype, Google Talk

or MSN Messenger), can be considered an indirect form of social net-

works, because they are naturally modeled as communications between

4 SOCIAL NETWORK DATA AN ALYTICS

different actors. One advantage of such applications is that the traces

of the communication data are often available (subject to some privacy

controls). This data can be used for extensive analysis of such social

networks. Some examples are the extensive analysis on the Enron email

data set [28], or the recent veri¿cation of the six degrees of separation

conjecture in the context of the MSN messenger data in [16].

In recent years, a number of sites have arisen explicitly in order to model

the interactions between different actors. Some examples of such social

networks are Facebook, MySpace,orLinkedIn. In addition, sites which

are used for sharing online media content, such as Flickr, Youtube or

delicious, can also be considered indirect forms of social networks, be-

cause they allow an extensive level of user interaction. In these cases,

the interaction is centered around a speci¿c service such as content-

sharing; yet many fundamental principles of social networking apply.

We note that such social networks are extremely rich, in that they con-

tain a tremendous amount of content such as text, images, audio or video.

Such content can be leveraged for a wide variety of purposes. In partic-

ular, the interaction between the links and content has provided impetus

to a wide variety of mining applications. In addition, social media out-

lets provide a number of unique ways for users to interact with one an-

other such as posting blogs, or tagging each other’s images. Which such

forms of interaction are indirect, they provide rich content-based knowl-

edge which can be exploited for mining purposes. In recent years, it has

even become possible to integrate real-time sensor-based content into

dynamic social networks. This is because of the development of sensors,

accelerometers, mobile devices and other GPS-enabled devices, which

can be used in a social setting for providing a dynamic and interactive

experience.

Finally, a number of social networks can also be constructed from spe-

ci¿c kinds of interactions in different communities. A classical example

would be the scienti¿c community in which bibliographic networks can

be constructed from either co-authorship or citation data. These can

be used in conjunction with the content of the publications in order to

derive interesting trends and patterns about the underlying papers. We

note that much of the analysis for the ¿rst case above applies to this as

well, though a lot of data and content is available because of the way in

which such documents networks are archived. A number of document

collections and bibliographic networks are archived explicitly, and they

can be used in conjunction with more principled data-centric techniques,

because of the content which is available along with such networks.

剩余517页未读，继续阅读

huchunpeng

粉丝: 2
资源: 23

《Social Network Data Analytics》：社会学视角下的信息技术算法详解

Hibernate 离线的配置方法(hibernate-mapping-3.0.dtd)

hibernate-mapping参数详解

hibernate mapping 文档

hibernate-configuration-3.0.dtd、hibernate-mapping-3.0.dtd

hibernate-mapping

hibernate-mapping中not-null什么意思

最新资源