Vanity or Privacy? Social Media as a Facilitator of Privacy and Trust

Jessica Staddon

Abstract

Google

In this position paper, we argue that social media provides valuable support for the perception of one’s self and others, and in doing so, supports privacy. In addition we suggest that engagement, which reflects a certain degree of trust, can be facilitated by social information. We support our arguments with results from a recent privacy survey and a study of social annotations in search.

1600 Amphitheatre Pkwy. Mountain View, CA 94043 USA [email protected]

Author Keywords Social; vanity search; social search annotations; privacy; trust

ACM Classification Keywords K.4.1.f Privacy

General Terms Human Factors; Privacy; Security

Introduction

Copyright is held by the author/owner(s). CSCW’12, February 11–15, 2012, Seattle, Washington, USA. ACM 978-1-4503-1051-2/12/02.

The abundance of communication that social media enables clearly can lead to privacy problems, often with severe personal consequences. Jobs have been lost, marriages ended and court cases won all because of unintended sharing of online social communication (e.g. [5, 7]). This online social communication often comes with two

types of audiences. First, there may be a specific target for the communication (e.g. the owner of the Facebook wall on which a comment is left). In addition, there is often a broader audience who can observe the communication (e.g. in the case of a public Facebook wall). While this broader audience can lead to privacy mistakes (indeed it is essentially at the root of the privacy problems mentioned above), we argue it also leads to huge privacy advantages by facilitating perception, both in terms of understanding of one’s online self, particularly as driven by the inputs of others, and self-representation.

in doing so enables better privacy. In particular, we present survey evidence that “vanity” searches are associated with an important privacy need. We also present evidence compatible with the conjecture that social annotations in search support privacy by enabling better self-representation and thus more privacy-aware sharing. While these examples offer no judgment on whether social media is good for privacy in any absolute sense, they do support our contention that it is possible to design social media systems that are engaging and supportive of privacy and trust.

The former involves understanding how others perceive you, and is clearly a privacy win. External perception is key to managing reputation and trust. Understanding how others perceive themselves (self-representation) is also important to privacy, but less directly. Visibility into someone’s self-perception enables another individual to interact with them with far less privacy risk. If Alice knows something about Bob’s preferences, habits and overall persona, she is far less likely to approach Bob under the mistaken impression of a shared interest or hobby, compromising her own privacy in the process.

Vanity searches also serve privacy

This same visibility into social views/preferences can also impact trust, a concept closely related to privacy [2]. If a user has an affinity with those in an online community, they will likely feel more trusting of the community [3, 4, 6]. We argue here that social media can facilitate trust and engagement by reflecting the views of the community in aggregate.

The entertainment aspect of vanity searches is clear, yet they are also a valuable tool for understanding online reputation. While the increased usage of search personalization makes it less clear what the results will be when a particular person searches for an individual, a comprehensive review of the results returned for a given name provides a pretty good sense of the individual’s online representation. Indeed, results from a recent survey we conducted indicate that vanity searches are often closely associated with reputation

In this short paper, we present 2 examples of how social media enhances both aspects of perception, and

Vanity searching (or “egosurfing” [1]) is the popular practice of searching for one’s own name using a search engine. As the name suggests, it is generally viewed as a self-esteem promoting activity; an easy way to get a sense of what others think of you and what you have done (at least those activities with web representation). Social media has made the results even more rich with social networking profiles and posts potentially appearing along with self-maintained blogs and web pages and public records of transactions.

concern. This survey was conducted with 200 users randomly selected from a broad pool of paid study participants, the majority of whom have college degrees, are within 24-45 years of age and are slightly more likely to be male than female.

case they are Facebook “like”s. These endorsements may help the viewer to prioritize the search results. We also propose that these endorsements may help build trust between the viewer and the endorser, since they provide information about the endorsers’ preferences and interests. Indeed, these annotations may serve as a form of asynchronous, unilateral, social support that can help the viewer reduce privacy risk and transition to a more active member of an online social community. In support of these conjectures we present some results from studies of 615 users drawn at random from the same population as the survey of section 2. In each study, each participant received the title, snippet and url for each of several articles (seven, eight, nine and five in the four studies, respectively). The articles were all news articles and had appeared within the last couple of days prior to each study. The articles were taken from well-known online news venues such The Wall Street Journal, The New York Times and CNN.com.

Figure 1: Reported reputation concerns and reported vanity search behavior. Almost 50% of the users who reported they did not vanity search, were not at all concerned about online reputation.

Social Search Annotations and Trust The Google and Bing search engines now support the annotation of search results with endorsements from the viewer’s contacts. In Google’s case, the endorsements are in the form of “+1”s, and in Bing’s

Each study consisted of three conditions that differed according to the type of annotation used. In particular, each participant received an article title, article snippet and either a Facebook like count annotation a qualitative popularity annotation (e.g. “this article is one of the most popular on nytimes.com”) or no annotation, depending on the condition. We emphasize that the studies put the annotation and the article essentially on equal footing in terms of visual emphasis, in contrast to most publisher sites, where typically the article receives greater emphasis. The participants were asked about their interest in sharing the articles (among other questions). There

were an average of 95.33 users in each study condition, with very little overlap between users across studies.

articles with popularity annotations (3.32), versus Facebook annotations (3.23) and no annotations (3.17). We summarize the interest ratings of low sharers in Figure 2.

We found significant increases in engagement when annotations of either type were present. In particular, an analysis of variance (ANOVA), conditioning on annotation type, found significance at the .05 level for article and topic interest, and significance at the .01 level for bookmarking interest (no significance for sharing). We summarize these results in Table 1. Scale = 1-5 (5 is most interest)

Article Interest

Facebook Annotations (e.g. “this article has been liked on Facebook 643 times”)

No Annotation

3.22

3.21

Qualitative Popularity Annotations (e.g. this article is one of the most viewed on CNN online) 3.28

3.4

3.45

2.09

2.15

2.32

2.35

Topic 3.44 Interest Bookmarking 2.19 Interest Sharing 2.39 Interest Table 1: Interest responses

Figure 2: Users who share less than once a week and average article interest, topic interest, bookmarking interest and sharing interest, by annotation type.

While clearly more research is needed to fully understand the impact of annotations on trust an engagement (particularly research that more closely matches the settings of today’s online services) these results suggest that social information, even in aggregate, can encourage engagement which is likely associated with trust.

over all users

The 307 users with low sharing habits (defined here as those who share online content less than once a week) reported a preference for the popularity annotations over the Facebook annotations that is weakly significant (p < .1) in that their average interest was highest for

Conclusion We have argued that social media both facilitates reputation monitoring and trust building, both of which are compelling for privacy. A core challenge in realizing

these privacy benefits without compromising privacy in other respects is awareness. The content an individual provides that drives their self-representation must be provided with their awareness and understanding of how the aggregation of data contributes as a whole to their online representation. This is particularly challenging for personalized online services with which an individual may not immediately understand the perspective of others viewing them through personalized services.

References

1. G. Branwyn. Jargon Watch. Wired Magazine, March 2005. 2. X. Luo. Trust production and privacy concerns on the Internet: A framework based on relationship

marketing and social exchange theory. Industrial Marketing Management, 2002. 3. C. Ridings, D. Gefen and B. Arinze. Psychological Barriers: Lurker and Poster Motivation and Behavior in Online Communities. Communications of the Association for Information Systems, Volume 18, 2006. 4. C. Ridings, D. Gefen and B. Arinze. Some antecedents and effects of trust in virtual communities. Journal of Strategic Information Systems 11 (2002) 271–295. 5. Divorces blamed on Facebook. The Sun (UK). December 21, 2009. 6. M. Smith, C. Giraud-Carrier, N. Purser. Implicit affinity networks and social capital. Journal of Information Technology and Management, September, 2009. 7. C. Smith and C. Kanalley. Fired over Facebook: 13 Posts That Got People CANNED. Huffington Post. July

Social Media as a Facilitator of Privacy and Trust - Research at Google

In this position paper, we argue that social media .... one of the most popular on nytimes.com”) or no ... in contrast to most publisher sites, where typically the.

360KB Sizes 40 Downloads 444 Views

Recommend Documents

Vanity or Privacy? Social Media as a Facilitator of Privacy and Trust
each study, each participant received the title, snippet and url for each of several articles (seven, eight, nine and five in the four studies, respectively). The articles.

Glimmers: Resolving the Privacy/Trust Quagmire - Research at Google
matic Glimmer of Trust, which allows services to validate user contributions ... to much recent work exploring SGX in cloud services—to realize the ... cols; Hardware-based security protocols;. KEYWORDS ... what precisely they do with data they col

Trust 2.1 – Advancing the trust debate - Research at Google
May 1, 2007 - transactions to phishing). ✸ Any situation is embedded in a web of multiple trust relationships and risks. Trust 2.1 – Advancing the Trust Debate.

RAPPOR: Randomized Aggregatable Privacy ... - Research at Google
domain that motivated the development of RAPPOR: the need for .... example, if 75 out of 100 responses are “Yes” for a single client in the ...... No free lunch in.

CSCW12-Privacy and Social Media Ecologies.pdf
individuals in social networking sites engage in active ... An ecological perspective [5] provides a foundation to ... Gibson [5] introduced the concept of.

DownloadPDF Online Privacy and Social Media
DownloadPDF Online Privacy and Social Media (Privacy in the Online. World) FULL MOBI. Book Synopsis. Governments, businesses, and individuals all rely.

Rényi Differential Privacy - Research at Google
... differential privacy, while additionally allowing tighter analysis of composite hetero- .... Kairouz et al. give a procedure for computing an optimal k-fold composition of .... R . We observe that. Dα(PQ) ≥ Dα(g(P)g(Q)) by the analogue of the

Would a privacy fundamentalist sell their DNA ... - Research at Google
Jul 9, 2014 - A white paper from Google reports that GCS performed favorably against both a probability ..... blue), and the top right (most concerned, colored red) of each panel. The data-driven segmentation is ...... alerts social services that the

Would a privacy fundamentalist sell their DNA ... - Research at Google
Jul 9, 2014 - ios. If Westin categories were significantly correlated with scenario responses, we would expect substantial divergence between the means ...

Undo and Erase Events as Indicators of ... - Research at Google
Permission to make digital or hard copies of all or part of this work for personal or .... Seven laptops were identically configured with SketchUp. Participants ...

Comparing the use of Social Networking and ... - Research at Google
ture is consistent with the observed uses of microblogging tools such as Twitter and Facebook, since microblogging is commonly used to announce casual or daily activities [6]. As can be seen in Figure 6, Creek Watch users more com- monly (60%) post o

Perception and Understanding of Social ... - Research at Google
May 17, 2013 - media sites are making investments based on the assumption that these social signals .... They found that the most popular question types were rhetorical ..... ized social search results, 8-10 tasks for each subject. • During the ...

Social Media as a Communication Strategy.pdf
Social Media as a Communication Strategy.pdf. Social Media as a Communication Strategy.pdf. Open. Extract. Open with. Sign In. Main menu. Displaying Social ...

Xenophobia-social-media-research-public.pdf
Thus we will refer to non-media conversation. originating in South Africa, pertaining to xenophobia in South Africa as total. relevant conversation. Page 3 of 31. Xenophobia-social-media-research-public.pdf. Xenophobia-social-media-research-public.pd

Collaborative Human Computation as a Means ... - Research at Google
human computation performed by a user's social network .... cial Network and Relationship Finder (SNARF). ... [10] describe the vocabulary problem faced.

Large-scale Privacy Protection in Google Street ... - Research at Google
false positives by incorporating domain-specific informa- tion not available to the ... cation allows users to effectively search and find specific points of interest ...

Privacy Mediators: Helping IoT Cross the Chasm - Research at Google
Feb 26, 2016 - interposes a locally-controlled software component called a privacy mediator on .... be determined by user policy and access to this storage is managed by .... ing policies defined per sensor rather than per app results in fewer ...

PseudoID: Enhancing Privacy for Federated Login - Research at Google
is backward-compatible with a popular federated login system named OpenID. ... Although most identity providers will prompt users whether they want to reveal this information, identity providers ... For example, Plaxo, a social networking and address