Hi Morad, There are two engines that can be used for Sentiment Analyses:
* Sentiment WordClassifier Engine: This engine checks Tokens in the Text against a Vocabulary that defines sentiments. So after this Engine you will have sentiments assigned for single words. * Sentiment Summarization Engine: This engine consumes word level sentiments as well as POS tags and summarizes them for Phrases, Sentences and the whole document. For summarization it does not use a dependency tree. However it uses POS tags to detect the Noun (or Pronoun) a adjective (with a sentiment) is assigned to. It also considers POS tags that indicate a negation. Note that for the Sentiment WordClassifier Engine you will also require the vocabularies with the sentiment values of Words. They are provided by the data/sentiment bundles [1]. Just check out the source, build them locally and install the bundle to your Stanbol launcher. Alternatively you can also download the vocabularies your self and add them to the /stanbol/datafiles directory of your Stanbol instance. For more detailed instructions pleas see the README.md files of those modules A typical chain configuration for Sentiment Analyses includes * langdetect * opennlp-sentence * opennlp-token * opennlp-pos * sentiment-wordclassifier * sentiment-summarization best Rupert [1] http://svn.apache.org/repos/asf/stanbol/trunk/data/sentiment/ On Fri, Nov 29, 2013 at 1:31 AM, Morad Larhriq <morad.larh...@googlemail.com> wrote: > Hello, > I am new with Stanbol ,i want to write a function to analyse the sentiment > of a sentence.some one have an idea how i can begin. > PS: i installed Stanbol with netbeans. > thanks. -- | Rupert Westenthaler rupert.westentha...@gmail.com | Bodenlehenstraße 11 ++43-699-11108907 | A-5500 Bischofshofen
