search

shekhargulati / 52-technologies-in-2016

Let's learn a new technology every week. A new technology blog every Sunday in 2016.
https://shekhargulati.com/
MIT License
7.2k stars 597 forks source link

Feedback: Week 3 Stanford CoreNLP blog #5

Open shekhargulati opened 8 years ago

shekhargulati commented 8 years ago

Please provide your feedback by posting a comment against this issue.

szahn commented 8 years ago

Great post. What if we could easily tell if any article is positive or negative? I love the idea of adding sentiment to articles so you could easily tell if its first, highly emotional/opinionated and second, what that emotional bias is. I can see Facebook adding sentiment analysis to the news feed.

tylerjharden commented 8 years ago

Interesting post, I wouldn't say great compared to the other reading content you put up, but I appreciate the time. I think what would have made it more valuable is if you dove into the Stanford NLP framework itself a little more. Explaining the annotations, showing some of the feedback and values you get back, etc.

Definitely a solid primer, however.

shekhargulati commented 8 years ago

@tylerjharden actually I was working on slick post that day but blog was not shaping up well. So, I decided to write about Stanford CoreNLP with few hours left in the day. I agree I didn't covered Stanford CoreNLP in detail. Will definitely clean it up and make it better. Thanks for honest feedback. Also, I would be more than happy to accept PR if you are willing to add few sections and details :). Thanks Again.

tylerjharden commented 8 years ago

@shekhargulati If I knew anything about it, I'd be more than happy to. If I get the time I will gladly contribute, I like your model here.

s6o commented 8 years ago

Interestingly changing "Dhoni" to "John" changes the sentiment from NEGATIVE to NEUTRAL The sentence in the test: "Dhoni laments bowling, fielding errors in series loss" I'm not a native english speaker so perhaps I'm missing something here ...

mariussoutier commented 8 years ago

Too bad sentient analysis only works for English.