Social Computing Session 4: Social knowledge production and services

Online social recommendation systems vs. real world advice seeking

For a project for another class I have been locating materials about Geogiana, the Duchess of Devonshire. Before I complete the project I wanted to know what others thought about some of the items I have located. To that end, and for exemplars for this posting, I visited Amazon.com and BN.com to check out the customer reviews of The Duchess by Amanda Foreman. See the screenshots above. Amazon.com had more reviews available than BN.com though many of them were more like plot synopses or vague "liked it," "loved it," or "hated it" type commentary. There were a few that I found helpful with well-written, thoughful discussions with accurate spelling. I found the Detailed Ratings on BN.com very helpful because they really fleshed out what elements specifically the reviewer rated highly or poorly. Unfortunately there was only one rating/review to work with there. Amazon.com does include a tagging feature that may prove helpful as I continue to look for additional materials.

I found the article Social Annotations in Digital Library Collections particularly relevant in this case as the ratings and reviews function as a history of others responses to and interests in the text. While they are not page or even chapter specific, these recommendations can help a user decide whether or not an item is relevant and/or worthwhile for whatever purpose. Also, both Amazon and BN offer "people who bought this also bought" features that may prove helpful in finding additional related materials. Kristina Lerman's article about filtering also applies to some degree...BN.com provides both collaborative and social filtering elements for users of MyB&N. Unfortunately, my research in this area has been very recent and the system is designed to filter based on my overall use and therefore has not provided any additional recommendations for me. We'll see how long it takes the system to catch up with my new interest...if it does at all.

The trust issues discussed in previous sessions came to the fore while I was looking for reviews...how does one determine whether or not a review is trustworthy? This is especially true when one is looking for research materials.

So, in comparison, and also in search of a review I can trust, I spoke with a writer friend who recently began researching Georgiana for her book. I've known her for years and trust her ability to critically analyze material. Her comments about this book were in line with some of the reviews available online, but since I know her, her background, and her abilities I am more inclined to view her review and reviews that are similar more favorably. I also emailed an old history professor whose research included the era of Georgiana, though I have not received any details back yet.

Overall, I found this week's readings varied and interesting. I am particularly excited aboutthe potential future of libraries as they continue to adapt to mobile communications and the changing social needs and ideas. If our library database incorporated "annotations," or reviews by registered users, tagging, or other social computing and knowledge-sharing elements I would be ecstatic! I personally find research rather lonely and sometimes do not trust my own analysis...having a social history in this environment would make for more interesting research. At the moment, one can accomplish some connection by using bibliographic references, who has used what in what context and when...but the speed and variety allowed by the elements discussed in this week's readings allow for many more connections and discussions. As always, quality is a concern when allowing ratings, reviews, tagging, and other "unregulated" social elements, but if the system incorporated elements of social capital allowing these elements and the users themselves to be rated and reviewed, then perhaps the quality and trustworthiness could be analyzed with some degree of reliability.

Session 4 Citations

Allen, Stuart M., Gualtiero Colombo, Roger M. Whitaker (2009). Forming Social Networks of Trust to Incentivize Cooperation. Proceedings of the 42nd Hawaii International Conference on System Sciences, Waikoloa, HI, 5-8 January 2009.

Ellison, N.B., C. Steinfield and C. Lampe (2007). The Benefits of Facebook "Friends:" Social Capital and College Students' Use of Online Social Network Sites. Journal of Computer-Mediated Communication, 12(4). http://jcmc.indiana.edu/vol12/issue4/ellison.html

Eryilmaz, Evren, Mitch Cochran and Sumonta Kasemvilas (2009). Establishing Trust Management in an Open Source Collaborative Information Repository: An Emergency Response Information System Case Study. Proceedings of the 42nd Hawaii International Conference on System Sciences, Waikoloa, HI, 5-8 January 2009.

Gleave, Eric, Howard T. Welser, Thomas M. Lento and Marc A. Smith (2009). A Conceptual and Operational Definition of ‘Social Role’ in Online Community. Proceedings of the 42nd Hawaii International Conference on System Sciences, Waikoloa, HI, 5-8 January 2009.

Massa, Paolo (2006). A Survey of Trust Use and Modeling in Current Real Systems. Trust in E-services: Technologies, Practices and Challenges. Idea Group. http://www.gnuband.org/files/papers/survey_of_trust_use_and_modeling_in_current_real_systems_paolo_massa.pdf

Williams, D. (2006). On and Off the 'Net: Scales for Social Capital in an Online Era. Journal of Computer-Mediated Communication, 11(2), article 11. http://jcmc.indiana.edu/vol11/issue2/williams.html

Social Computing Session 4: Social Role, Capital, and Trust

There was a lot of material in these readings...to discuss I will be comparing and contrasting features from the two sites I chose, Citizendium (http://en.citizendium.org), hereafter refered to as CZ, and Care2 (http://www.care2.com).

The first concept discussed in our readings is that of "social role," in A Conceptual and Operational Definition of 'Social Role' in Online Community. People play many roles in their lives, different roles for different environments, situations, purposes. I found the discussion of structure and culture very interesting in this reading. There are roles that cross cultures, that are identifiable wherever one goes. I found the idea of roles in the online community intriguing as I hadn't really thought about roles online much as I base most of my social conceptions on real world interactions. I did find, however, that some sites have very clearly defined roles, while in others the structure is less identifiable. Take CZ as an example of the former. CZ is a reference site similar to the well known Wikipedia. One notable difference is that CZ requires that members be approved before they can begin to contribute to the site rather than being an end-user and the site has three very clearly defined roles. The authors are the majority of registered users, and are the people responsible for proposing and creating content. Editors are authority figures responsible for maintaining consistency and performing other editorial functions on the work of the authors. There is a second authoritarian role in CZ, the Constabulary (see screenshot above). These individuals oversee site policies and are a behavioral rather then editorial authority. In Care2, roles are less well-defined and more social-based, though one can proclaim one's activist status in terms such as "rabid" or "casual" in one's profile. One would assume that ones's actions would then need to back up that proclamation if one expects to gain and maintain social capital on the site.

Social capital...this concept is fascinating in that I found it astonishing that this could be measured to any satisfying degree. Color me surprised. I also found that the article, On and Off the 'Net: Scales for Social Capital in an Online Era, did a good job of identifying the positive and negative perspectives and also in labeling "bridging" and "bonding" as distinct types of social capital. It seems to me that both CZ and Care2 are both primarily involved in "bridging" in that they bring together people from all over to contribute and communicate about a particular topic or interest. Care2 offers some "bonding" characteristics in that I know my use of the site was instigated a little over a year ago in support of a petition put together by friends. There is also a singles area, where the site may act in part as a dating service, I don't know for sure because I am not registered with the singles area, but the description seems to indicate that it is a place to meet people whose values mesh, possibly leading to romantic relationships.

Trust is a big deal to me, especially when online, so I found these readings extremely valuable. Paolo Massa defines trust as "the explicit opinion expressed by a user about another user regarding the perceived quality of a certain characteristic of this user." The emergency management article didn't quite strike a chord for me, but the third trust related article, Forming Social Networks of Trust to Incentivize Cooperation certainly did. In particular I found the idea of reputation satisfying. CZ clearly attempts to build trust with it's users from the get-go, by insisting that producers of content use their real names and be approved by the site before they can contribute. The use of real names draws a clear connection between reputation in the virtual world and the real world. CZ also has clear avenues of enforcement of value and behavior in it's two authoritarian roles, the Editors and Constables. Care2 has multiple incentive mechanisms to build trust and reputation. These include "green stars," "butterflies," "testimonials," and "golden notes." These are all incentives posted to one's profile, either by fellow individual users (green stars, testimonials), the community (golden notes), or the site itself (butterflies - a record of one's activity on the site). The user and community generated incentives have clear social capital and trust worth, indicating that others in the community recognize and approve of a particular user. The butterflies, as automatically generated records of one's actions in particular areas within the site, are a true representation of a user's actions. Interestingly, the stars and butterflies are recorded in total permanently, but the icons actually "disappear" from a profile after a month. I found this an ingenious device for driving continued participation! By collecting stars and butterflies, testimonials and golden notes, a Care2 user has several incentive based tools to stay active and build a reputation on the site.

Overall I found this session enlightening and may include concepts discussed here in the final project. I am particularly interested in young (teen/college age) users and their perceptions of privacy and trust in their use of online resources.