MeeHive is a new service that seeks to offer users a personalized newspaper, made up of stories from sources, including a large number of blogs and news sites. The site harnesses the power of Kosmix, a universal search engine that pulls information from various sources to produce comprehensive thematic pages.
When a new user enters the site, they are asked to enter some of their favorite topics, which can range from main lines such as "Sports" or "Technology" to more specific areas, such as "Stem cells". Users can also specify certain companies or keywords that they want to track. Based on this, MeeHive creates a digital newspaper with content from various sources, including news sites like CNN and a wide selection of blogs (the system uses an authority algorithm to choose the best content). The Kosmix algorithm works well, extracting the relevant stories without merging errors (at least on common topics) and ensuring that the same story does not appear repeated multiple times.
The interface of the site is well done, presenting a large amount of information but without the user being overwhelmed by an excess of information. In addition to the collection of stories, classified by topic, the site also displays the most recent Tweets related to the user's favorite topics. Unfortunately, at the moment the interface of the page is static, although a future update will allow users to rearrange their panels however they want. In addition to the Web edition, users can choose to receive daily summaries of their newspapers via email, and view them on the site's iPhone application.
Overall MeeHive works quite well, but will people use it? Custom news sites are not new; It is an idea that has been put into practice many times and, in general, it has not gone so well. But, unlike some of these initiatives, MeeHive also incorporates a social element in its site, that of allowing its users to share the news they like with their friends, something that can highlight the most interesting stories better than any algorithm.
Source: Science Daily