• Guest

    Maybe you should include your name in the description? 🙂

  • ffg jhjhkj

    do you have a name or someone hacked it???

  • Pingback: #StopRape debate – what was the sentiment of the Twitter campaign? [interactive] | Grubstreet()

  • Praveen Gr

    I am impressed with your online demo. can you help me with sentiment analysis code or guide with some useful links ?

  • asd


  • dads


  • Allen Bruns

    Hi Jacob, Do you know some NLTK functionality of parsing name and US postal address out of text? I am using REGEX, and its brutal in regards to varying text formats.
    Thank You!!

  • Hi Allen, I don’t know if any support for that in NLTK. But maybe take a look at https://github.com/datamade/usaddress. Regex is definitely not going to be good enough. You may also want to consider trying to train a tagger and/or chunker on addresses, meaning that you could use NLTK if you had training data that treated addresses as named entities that you’re trying to extract. Hope that helps.

  • Marian Bratu

    Hi Jackob! I’m having a problem with the MongoDb corpus example from your book, when having two projection fields in MongoDb client. Could you please give me a help somehow ?(give me, please your e-mail, or whatever way you want, in order to explain more detailed)

  • The code for that example is old now, and I wouldn’t be surprised if things have changed for Mongo that break things. I can’t really support it at this point, and it sounds like a MongoDB specific question, so I’d recommend asking on their users list or support forum.