Peer reviewing the Internet? Crowdsourcing online identity and reputation

We have all heard at one point that you cannot believe everything you read.  This is especially true when it comes to content on the Internet.  While we feel some sources are more reputable than others and therefore the content should be more trustworthy, in reality, a good reputation does not automatically equal the truth. It is hard for the general population to suss out whether the information they are reading is actually good or bad. Some of us might be skeptics or interested enough in being well informed to fact check many sources in order to verify the truthfulness of the information we are seeking. On the other hand, I would say many people simply want to believe what they read, especially if, in their opinion, the source seems credible.

This leads us to an interesting conundrum regarding the concepts of reputation and identity. While reputation might be the focus in the public’s eye, the foundation of that reputation comes from its identity. And the identity is what must be accurate ultimately.

A new project entitled Hypothes.is attempts to address this problem. According to its Wikipedia page, Hypothes.is is an “open-source software project that aims to collect comments about statements made in any web-accessible content, and filter and rank those comments to assess each statement’s credibility.” In a nutshell, the project seeks to peer review the entire web. The project, spearheaded by Dan Whaley (co-founder of one of the earliest online travel booking sites), has a Kickstarter page in which it has already reached its $100,000 goal in order to develop a prototype behind the idea.

The project developed out of the issue that reputation and identity are two separate concepts. There is simply no system in place that is unbiased. Popular social media sites such as Facebook and Twitter demonstrate the importance of popularity among the public. However, sites like these are not able to demonstrate that people have many facets, and may want to express one side in one context but another side in another. Right now, it seems a main identity such as a Facebook identity, while loosely tied to or extended to other identities, still influences other identities and therefore the reputation the user may have. Essentially, the bias from these popularity sites provides an inaccurate identity.

Hypothes.is aims for a practical system that shows a person’s many facets in their identity, compared to the current sites that rely on spreading one constant, unchanging identity for everyone. The reason this is dangerous is that only a small percentage of people can protect their public identity and reputation due to resources.  However, for most of the general population, the ability to express yourself freely on the web may cause someone to be marginalized or vulnerable for doing so. Hypothes.is wants to provide everyone the chance to participate equally and without hesitation.

Hypothes.is held a workshop recently and determined that an expensive, pseudonymous online identity would work best. A pseudonym would be easy to obtain but hard to change or create a new one. The user could be privately verified in some way so that the reputation system would know the true user but the web would know the pseudonym. Also, the pseudonym system could work in conjunction with other identity services (i.e. Facebook) if the user chooses to have these identities linked. I suppose the expensive requirement is to ensure people understand they will only obtain one pseudonym.

The project will apparently have its own reputation algorithms, mechanisms and moderation strategies. The annotators will most likely be trusted people from all over the world, similar to editors or journalists, and probably hold some sort of expertise. The annotations will then have to be displayed on the sites in a consistent interface in order for the project to be successful. = Initial concerns are how to keep the system objective when it is based on the work of subjective people. People are susceptible to bias and disputes and there will have to be another system in place to resolve differences between annotators. I also think the interface must be easily accessible to the user in order for the project to be effective. If it is too convoluted, I feel the user would just move on to another site, disregarding any information imparted.

It will be interesting to see the initial prototype of the project and how it will be implemented.  In the meantime, progress can be tracked at its Kickstarter page.