web2express.org

September 25, 2009

Following topics made easy

Filed under: news, social media, twitter — aj @ 12:55 am

If you don’t have a twitter account, you can still follow topics of twitter conversations. This new feature was introduced to web2express digest website yesterday.

Web2express digest application uses NLP and semantic technologies to automatically discover new topics in twitter conversations and fresh blogs in real time. The discovered topics are listed and searchable for the public. Users can select any topic to view the related tweets. This function does not require login.

While viewing topic tweets, any user can post tweets after logging into digest application. If the user provides a valid twitter account in his/her profile, then the user’s twitter account is updated. If the user does not have a twitter account, web2express will use the default @web2x account to post tweets. However, the tweets are sorted so that the user can see his or her tweets immediately.

Hope this new feature will make it easy for people without twitter account to participate in public conversations.

September 20, 2009

Web2express Digest Now Analyzing Millions of Tweets/Blogs in Realtime

Filed under: news, social media, twitter, web intelligence — aj @ 5:06 pm

A huge data stream of blogs has been added to the web2express digest server for a couple of weeks now, but the front end http://web2express.org was finally connected to this important addition of real time data yesterday.

For more than half a year the system has consumed only twitter streaming data, which is about 1000 tweets per minute.  New topics are auto-discovered from these twitter conversations every minute as popular topics emerge.

The blogsphere is another rich source of fresh conversations. It’s much bigger than twitter stream in bytes. Thanks to Kevin Burton, CEO of Spinn3r, I am permitted to pull in Spinn3r’s blog data stream to my real time analysis system.  We met on the event that I organized for twitter applications 2 months ago. Kevin was so generous to suggest me to try their data service. Spinn3r crawls the whole web for blogs and then make the data available through API. It’s not exactly in real time, but very close. Roughly one million blogs per 24-h period (after filtering) is now flowing from spinn3r to web2xpress.

Combining both twitter conversation data and spinn3r blog data, web2express digest server now analyzes 2-3 million fresh conversations and blogs on a continuous 24-hour basis. The primary analysis is still auto-discovery of new topics in real time conversations. Immediately after new topics are auto-discovered, they are provided to the frond-end web page, allowing users to follow realtime twitter conversations and blogs by topic.

It’s exciting to see such a huge amount of fresh web contents being analyzed in real time! I’m pretty sure this real time system can have many different applications or usages. For example, advertisers and marketers can use it to monitor products and brands on social media, do sentiment analysis, and identify potential targets. I’m interested in hearing feedback from marketers and finding ways to make this real time tool useful for social media marketing.

Powered by WordPress