Download to read offline
![{
“entities”: [
{
“entity”: “2018”,
“type”: “year”
}
],
“intents”: [
{
“intent”: “GetSickDays”,
“score”: 0.99853384
},
{
“intent”: “None”,
“score”: 0.07289317
},
{
“intent”: “GetVacationDays”,
“score”: 0.0167122427
},
{
“intent”: “GetPersonalDays”,
“score”: 1.0919299E-06
}
]
}
“How many sick
days do I have in
2018?”
Language Understanding Models](https://image.slidesharecdn.com/caapchatbotsnodejs-170821144742/75/Leveling-up-your-bot-with-language-intelligence-13-2048.jpg)
Leveling up your bot with language intelligence!
- 1. Leveling up yourbot with language intelligence! // Kevin Leung - @KSLHacks // Technical Evangelist, Commercial Software Engineering // Microsoft
- 2. Why Conversation asa Platform? “Bots are the new apps,” said Nadella during a nearly three-hour keynote here that sketched a vision for the way humans will interact with machines. “People-to- people conversations, people-to-digital assistants, people-to-bots and even digital assistants-to-bots. That’s the world you’re going to get to see in the years to come.” - USA TODAY “Microsoft CEO Satya Nadella says future belongs to bots”
- 3. QnA Maker Did youeven check the FaQ before asking me?
- 4. Microsoft Bot Framework+Artificial Intelligence
- 5. Microsoft Bot Framework SupportsNodeJS and C# .NET Bot Emulator provided for local development/testing Easy to incorporate Natural Language Processing (Cognitive Services) Easily deploy to Azure Cloud Services Lead users through conversation Prompts, Choices, Media, Rich Cards (Displays)
- 6.
- 7. Bot Connector ■ Connectorto Platform Channels – One Bot : Many Channels and endpoints ■ Why restrict yourself to one platform? – Hit them all ■ Session data on the user – remember who they are
- 8. Typical Bot ArchitectureDiagram
- 9. Creating a Bot Livecoding.. What could possibly go wrong
- 10. Adding intelligence and knowledgebase Create, train, deploy NLP models with ease.
- 11. LUIS: Cognitive Services Cognitive Services: LUIS abstracts this into a training model Machine learning and deep learning algorithms abstracted Using artificial intelligence to process natural language Extract Intents (meaning of the utterance or phrase) Extract Entities (items in the utterance that are of value)
- 12. LUIS.ai Utterance: ‘I want tolearn a lot and becoming inspired with Breaking Into Bots at Midwest JS!’ Entities: sessionTitle | conferenceTitle Intent: becomeInspired
- 13. { “entities”: [ { “entity”: “2018”, “type”:“year” } ], “intents”: [ { “intent”: “GetSickDays”, “score”: 0.99853384 }, { “intent”: “None”, “score”: 0.07289317 }, { “intent”: “GetVacationDays”, “score”: 0.0167122427 }, { “intent”: “GetPersonalDays”, “score”: 1.0919299E-06 } ] } “How many sick days do I have in 2018?” Language Understanding Models
- 14. Reduce labeling effortwith interactive featuring Seamless integration to Speech API Deploy using just a few examples with active learning Supports 5 languages (English, Chinese, Italian, French, Spanish) Language Understanding Models
- 15. LUIS Demo Create andtrain a language model
- 16. Case Study Empowering yourbusiness with chatbots & AI
- 17. Instaply bridges thegap between what your customers want, and what you’re currently offering (call, email, live chat), by creating a text/messaging channel that will help you improve customer experience and drive sales.
- 18.
- 19.
- 22. More on CognitiveService and IoT @ 14:30 (Auditorium)
- 23. // Kevin Leung- @KSLHacks // Technical Evangelist, Commercial Software Engineering // Microsoft ‘resources’: { dev.botframework.com LUIS.ai microsoft.com/cognitive github.com/KSLHacks } ‘realStories’: { aka.ms/techcase }
- 24. // Kevin Leung- @KSLHacks // Technical Evangelist, Commercial Software Engineering // Microsoft ‘resources’: { dev.botframework.com LUIS.ai microsoft.com/cognitive github.com/KSLHacks } ‘realStories’: { aka.ms/techcase }