What's Love Got To Do With It?...
summary
Claim: language shapes thought and automated sentiment needs concrete data checks. A Rails WhatsApp app fetches headlines from NewsAPI, sends them to IBM Watson NLU for sentiment and emotion scoring, and replies via Nexmo Message API using JWT auth and Ruby gems. The build runs and returns surprising specifics: only 38.8% of headlines match the topic, emotion shows ~55% joy and ~46% anger while sentiment is −0.62. The talk emphasizes tool choices, sorting by popularity, and validating the returned data.
AI-written summary · about
chapters (36)
- 0:00Introduction
- 0:57Who am I
- 1:50Agenda
- 3:10Why does this matter
- 3:32The big claim
- 4:57Hebrew
- 8:36Understanding Language
- 9:08Natural Language Understanding
- 9:21Turing Test
- 12:34Parts of Speech
- 13:11Polarity
- 13:53Broken Cars
- 14:14Sarcasm
- 15:18Contradictions
- 16:40Natural Language Processing
- 17:43Mood of the News
- 18:16Get her done
- 20:05Gem Files
- 20:33Routes
- 21:25Environment variables
- 22:10Controller
- 23:27Inbound
- 24:36Analyze Headlines
- 28:18Get News Headlines
- 29:44Send Message
- 30:24JWT Token
- 32:01Analyze your data
- 32:11Headlines
- 32:51Feminine
- 33:06Examine your data
- 33:26Data analysis
- 33:37Emotional Results
- 33:48Key Concepts
- 36:23Where Does This Leave Us
- 36:52Thank You
- 37:05Resources
related talks
30mWhat's Love Got To Do With It? Ruby and Sentiment Analysis
46mPractical Machine Learning and Rails
31mSyntax Isn't Everything: NLP For Rubyists
33mSyntax Isn't Everything: NLP for Rubyists
43mGetting Sentimental: exploration of iOS sentiment analysis
38mMachine Learning for Poets: The eMiLy App
10mThe future is here, and it started 5000 years ago
39mThe art and science of making computers understand sentiment and emotion