| Website | allvoices.com |
| Blog | allvoices.wordpre... |
| @allvoices | |
| Category | Consumer Web |
| Phone | (415)-391-1073 |
| Employees | 10 |
| Founded | 2/08 |
| Description | Global citizen journalism hub |
| Series A, 7/07 VantagePoint Venture Partners | $4.5M |
Allvoices (www.allvoices.com) uses technology and a global community to collect, validate and distribute user-generated news from around the world.
The site creates an almost instantaneous and fully-realized news package around as little information as a text message or tweet by pulling in and wrapping around related information from mainstream news and user generated content like blogs, videos, and pictures. Called the “Automated News Room,” Allvoices presents a 360 view of news events from multiple perspectives. The same technology vets contributions for authenticity and credibility, giving Allvoices citizen journalism’s first and only “Report Credibility Meter.”
In just over a year, Allvoices has attracted more than four million unique visitors a month and a world-wide community of contributors in excess of 200,000, making Allvoices the largest user-generated news site on the web.
Using proprietary algorithms, Allvoices scours more than 3,500 mainstream feeds and online, real-time sources of news like Twitter to provide relevancy and context for user-submitted news reports. This provides a level of security and vets contributed stories for authenticity. Story rank on the home page is determined by the number and diversity of other sources citing the particular news event, and community activity on the report.
The algorithm behind the Report Credibility meter is a proprietary, multi-faceted and adaptable approach to evaluating the trustworthiness of a report. At the core of the Allvoices’ system are the everyday citizens and contributing-journalists that create news reports. The Report Credibility meter is an evaluation of the actual report content, as well as the reputation of the author of the report. As stories move through the different pages on Allvoices, the meter registers new levels of credibility for both the story content and contributor. The final determination of credibility is evaluated on a number of criteria, including:
• How much does the community like or dislike content from the author? • How much does the community like or dislike the actual report? • How many other Allvoices contributors are reporting about the same thing? • How many other mainstream sources have supporting content about the report?
As users interact more and more with the Allvoices platform, the system learns more about the user, which in turn allows Allvoices to more accurately determine the authenticity of their reports. However, contributors to the site don’t have to be long-time users. A new user can also have high credibility for their contribution as a function of time. When a new user reports a story it enters the Allvoices system with low credibility. But, as the story moves through the site’s three news cycles—“This Just In,” “Breaking,” and “Popular”—it gains credibility, and will appear as a top-ranked news item. In turn, the story credibility helps with the reputation of the new user.
Either alone—just people or just the algorithm—would be insufficient as a way to vet stories, but Allvoices’ man-machine collaboration complements the strengths of each.
