Natural language processing fun

Posted in Zemanta|7 Comments

imageOur new demo gives a glimpse of what we are currently working on. For those that are following us that might not seem as big functionality advance since Odprti kop. However we have rewritten the whole backend, making it scalable and extendable. And in turn making us much more flexible.

Part of backend consists of Natural Language Processing – in short NLP. The fun thing is that no matter how many things you improve, there is always wast amount of input data that causes wrong or at least not-exactly-right contextualization. However I think we made a lot of important advances.

By the way: NLP in slovenian stands for Neznani Leteči Predmeti, or translated to english Unknown Flying Objects – UFOs. Which sometimes seems as quite an accurate description.

Subscribe to newsletter

  • Pingback: Natural Language Processing | Pixelturf

  • Pingback: Improving Automatic Translation Software, using Human text-corrections in Automatic Translation « OMADEON

  • Pingback: Zemanta ltd. » EuroPrix TopTalent Award

  • http://evandrake.blogspot.com Evan Drake

    Glad to hear that I’m the first English-speaking user. In fact, if you need anybody to work for you, I am a web developer with a journalism background, looking for side work.

    I attempted to use your service yesterday, but I was given the same picture for the second time in a row – even though I input completely different content. It was strange. I got a picture that said, “You got served.”

    I am glad to hear that you will soon be translating the system into English, and I would love to have a way to automatically blog a post from your service to Blogger. Perhaps you could use a Blogger API for this?

    I would expect the links to go out to random pages, but ones that would stick around for a significant amount of time. For example, I’d rather your system not create links to news articles, but instead to articles such as those in Wikipedia, as you are doing. In addition, I would forsee that your links would take into consideration the words that surround the particular link. I am certain that there are algorithms for such things, and I would be interested to hear about the technology behind this.

    Also, is there a way to create dynamic links that could change over time? In other words, one week the link for a word such as “lightning” might go to a page about the history of lightning, and another week it would go to a page about recent lightning strikes.

    In about a year or so, I plan to publish my blog postings in an online book. Adding your links and pictures to my writing is exactly what I was looking for. Perhaps your computer program will be a co-author for my book.

    If you’d like to reach me by email, it’s evancdrake (at) gmail (dot) com, and my phone number is xxx.xxx.xxxx.

  • http://www.zemanta.com andraz

    Oh Evan, please drop us a mail, we’d like to get in touch with you!

  • http://www.zemanta.com andraz

    Wow you are our first english-speaking user! Congratulations! We’d like to hear more from you.

    The answers:
    1. Yes, we definitely plan to offer solutions that are seamlessly integrated into other software
    2. Currently it only works for Slovenian language. If you’d written your posts in slovenian you would had have more links. English version is planned.
    What sites would you expect the links to?
    3. Yes, tags are going to get much better as soon as we move to english.

    There is still substantial work ahead of us and we are very eager to hear from future users what they want

  • http://evandrake.blogspot.com Evan Drake

    I like what you have going on here. I write a blog in a stream of consciousness style, and I’ve been looking for a way for a program to link the words that I wrote. This adds to the creative process for me, because now I have a computer interpreting my words. I am now using your demo interpretation on my site, after writing a post. Here’s what I want:

    1. The ability to automatically import your language processing into Blogger, along with the tags, pictures and links.
    2. More links in the content. Right now it seems to only give me one link, which is always to a page from Wikipedia. Is this intentional?
    3. Better tags. Some of your tags are in Slovensko, and I need them all in English.

    Thanks so much. This is wonderful!
    Evan Drake