Text Mining Text Mining / Items

Question about Named Entity Recognition

Get Feed
 Hi there,

I'm trying to tap into the knowledge pool of my fellow Twiners and maybe you can help.

I'm currently evaluating the usage of tools like Open Calais for one of our projects. The problem we have with most of the tools we found until now is the restriction of the supported languages, most of time English.
Do you know of any Name Entity Recognition/Metadata Extraction tool like Open Calais that supports at least English, Chinese and German (maybe even more languages)? Ideally, the license is a generous as Open Calais', free for commercial and non-commercial usage.

Looking forward to your suggestions,
Carsten

Comments

  • Public Comments

    • 20 months ago


      I posted this to the Text Mining Twine, but this Twine is more active (sorry for the repost)
      NLP
    • 20 months ago


      That's a tough set of requirements. Chinese is particularly hard to find. German might not be so difficult. But I don't know of a solution.
      NLP
    • 19 months ago


      HI,

      I´m sorry :-( , I´m using SAS text miner to do this kind of task, but it isn´t free. I wish you look! If you find something interesting about it, please share with us.

      cheers,
      Schiessl
      Text Mining
    • ben
      19 months ago


      Would agree with Nova on that! I know that some commercial tools have capability in Chinese...but they come with a pretty hefty price tag! The Encyclopedia Britanica engine and Lockheed Martin's Aerotext spring to mind as potentials to fulfill the language element.

      Sorry I can't be of more help. I will however mention your question at my next internal NLP meeting and see if anyone has got any good ideas/suggestions and revert to you.

      Good luck and let us know if you find something.

      Ben
      NLP
    • 19 months ago


      LingPipe is pretty popular for NE, and the license is free (for non commercial use).
      Text Mining
    • 19 months ago


      SAP, nee Busines Objects, nee Inxight's Entity Extraction package supports multiple languages but isn't free. Wikipedia lists this and other commercial and freeware packages (but does not list language support) here: http://en.wikipedia.org/wiki/Named_entity_recognition
      Text Mining
    • 19 months ago


      Hi, if you can put in some learning and developing effort you should take a look at GATE . It works for English out of the box, and a number of plugins exist that provide NER capabilities for Chinese, German, French, Romanian etc. Probably not the best coverage, but it's fairly easy to extend if you can pull together resources from somewhere. Plus you have complete control over why what is happening/being extracted. The downside (of course there is no free lunch..) of it is that you need to invest some development effort. However, everything is well documented and the community and support are nice folks.
      Text Mining
    • 19 months ago


      Thanks for your help! I will try some of your suggestions and report on the results here.
      Text Mining
    • 19 months ago


      Thanks for your help. If I find something, I'll let you know.
      NLP
    Add a Comment
Report This

Twine is about discovering, collecting and sharing the content that interests you. Learn More

Join Twine

Stats

First Posted By

First Comment By

Who's Interested In This?

Forgot your password?