Does Spam have an opinion?
summary
Claims that spam can be parsed for sentiment and topics and even turned against scammers using cloud ML, demonstrating Azure Cognitive Services Language API and the Text Analytics SDK with opinion mining enabled. Uses a real stat — 362 billion emails in a week with ~84% spam — feeds ~20 random junk messages through the service, shows a pie of categories, and runs a .NET console demo hitting the endpoint. Also cites the old Re:scam idea of chatbots wasting scammers time and notes the free Azure tier and practical success of extracting topics and sentiment.
AI-written summary · about
chapters (12)
- 0:00Start
- 6:30Predictive Modeling
- 7:21Spam Emails
- 10:02Azure Portal
- 10:57Types of Cognitive Services Categories
- 11:17Machine Learning Models
- 14:26Managed Identity
- 15:39Cognitive Services
- 20:59Opinion Mining
- 23:59Logging Opt Out
- 29:09Azure Ai Text Analytics Library
- 39:36Bing Speech Api
related talks
45mBuilding Emotionally Intelligent Apps using Azure AI
21mFast and Efficent Text Analysis using Azure's Language Cognitive Service
46mML.NET for developers without any AI experience
01h 05mSocial Text, Sentiment, and Tone Analysis
01h 09mDialogue with your Data: Interactive AI with Azure Cognitive Services
24mHighly Scalable Discriminative Spam Filtering
51mTaking Models to the Next Level with Azure Machine Learning
27mIs that spam in my ham? A novice's inquiry into classification.