Since I announced TextMinr, the Text mining as a service platform last week - the interest has been overwhelming: it got to the top 3 of Hackernews, it got retweeted by Tim O’Reilly of O’Reilly media and many other influential people in the tech scene.
Progress over the last week has been good, the public API is coming along and it’s quite likely we will start to dripfeed invitations to the beta towards the end of next week. I say “dripfeed”, because there are so many people who have declared interest that we’ll have to do it slowly to scale the service in a managed way as the load grows. Everyone who has signed up will get an invite (and people can still sign up), it just might take a week or two longer before everyone gets in.
Further, the first beta drop will likely be primarily a developer oriented release: we will provide API’s and documentation for:
- web scraping
- name/entity recognition
- subject identification
- context searching/extraction (extracting important words and sentences for a search term)
- document similarity (comparing two documents/urls for similarity of content)
- RSS listening that hooks into all of the above
The other functionality that has been mentioned, such as sentiment analysis, classification, Twitter and Facebook monitoring and analytics will be rolled out in the weeks thereafter (it’s a Beta after all - functionality will be continuously improved and added).
The first release may be mostly developer oriented, but in coming releases we’ll hopefully improve the polish of the web experience as well, so that non-developers will eventually also be able to create their own analysis pipelines and analytics reports.