How to extract text from web page?

Has anyone used NLTK to extract noun phrase embedded in web page content? How much work does it take to write a NLTK code for extracting noun phrase from web page? Also, do I need to train it in order to extract noun phrase from text in English. Can I use the default module without any training. Tha

• How would one compare NLTK to openNLP? I am a newbies to text analysis. I need to write a code for extracting and identifying noun phrase from web page. Any suggestion or input are appreciated. Thanks,

• Answer:

  For Noun Phrase extraction, I think OpenNLP is much easier to use than NLTK. Since you are starting with web pages, for you finding Noun Phrases will involve the following steps: crawling, HTML parsing, sentence tokenization, word tokenization, POS tagging, phrase chunking, and finally filtering out NP chunks.   OpenNLP comes with trained models to do this, so its just a matter of writing some boilerplate Java code.   With NLTK you will need to build your own POS tagger and phrase chunker out of smaller components (regex, backoff/classifier hybrids, etc). NLTK does come with its own HTML to text converter.