3 ms·
Show HN: Sentiment analysis of tweets
- amend 9y agoThese are four charts containing statistics about tweets from the Austin area. The charts discuss a sentiment analysis of the tweets and when the tweets were sent. Red means the tweet was catagorized as positive (for example: What a great day), blue as negative, and green as neutral. The last chart contains the 50 most common words used in the tweets. I think the results are pretty interesting to see, but if you're interested in a TLDR: People tweet sad things on weekdays and happy things on weekends, and tweet sad things late at night and happy things in the afternoon and evening. https://github.com/amend/austins-sweet-and-sour https://github.com/amend/austins-sweet-and-sour
- DrScump 9y agoYou might get more readership if you submit using a transparent domain address. Many people won't click on bare IP addresses.
- TekMol 9y agoStatistics without info about sample size and why the data is supposed to be significant and not just random noise are pretty much worthless. And the wordlist ... simply impossible. One of the 50 most common words is "auditorium". And "the" and "a" are not even in the list.