News for hackers. Stuff that matters.
Re:An Efficient Way to Extract the Main Topics from a Sentence by run4yourlives on Thursday May 09, 2013 @01:07PM (#5681576)
The only reason that sentence is "obvious" to many people is because we have a reference to a famous band from Liverpool that got its start in the late 50's/early 60's that is already embedded in our brain's library of facts.

Removed from that context human beings see that sentence as equally meaningless as a parser, because it is. I'd imagine many young people (who don't have the "correct" reference points) wouldn't have a clue that the sentence is about George Harrison.

In order to properly handle this sentence one would need the same external reference that your brain has. Without the reference the sentence can be discarded as incomplete, since that's what a human would do too.

Re:An Efficient Way to Extract the Main Topics from a Sentence by unhammer on Friday May 10, 2013 @06:40AM (#5685488)
You don't necessarily need knowledge to distinguish a topic from a subject though. It's a grammatical distinction. See

http://en.wikipedia.org/wiki/Topic%E2%80%93comment#Definitio... e.g. in the sentence (3) As for the little girl, the dog bit her "the dog" is the subject NP, "the little girl" is the topic. That toy example is perfectly parsable without semantics or even probabilities (though take any real-world sentence and I'm betting you'll need more than just grammar).

Re:An Efficient Way to Extract the Main Topics from a Sentence by teeja on Saturday May 11, 2013 @10:11PM (#5692848)
Thank you for elaborating my point. But a machine don't need an enormous library of facts to find the topic, just a small one. The words 'he', 'band', 'Liverpool' and '1958' are enough to get very precise ... but not because of their grammatical position.
Re:An Efficient Way to Extract the Main Topics from a Sentence by yolesaber on Thursday May 09, 2013 @04:23PM (#5682724)
This is why systems such as MIT's ConceptNet exist. NLTK and other NLP systems can provide the grammatical breakdown and ConceptNet can fill in the semantic / contextual gaps.