End-to-end dialogue systems, or a feature which wasn’t meant to happen | Rasa Summit

Rasa Technologies
Rasa TechnologiesRasa Technologies
End-to-end training, or a feature which
wasn’t meant to happen
Evgeniia (Genie) Razumovskaia
PhD on Computation, Cognition and Language, University of Cambridge
End-to-end training
● … as a new feature of Rasa
● … as an example of how research proceeds within Rasa
What’s on:
● What is end-to-end training?
● Why end-to-end? Why now?
● What was the research process?
● What happens after research? (my biggest surprise 😅)
What is end-to-end training?
Traditional pipeline
End-to-end training:
predict the response text directly from the preceding text
Image from here with little modifications
What is end-to-end training?
- story: end to end happy path
steps:
- user: “hi”
- bot: “hi!”
- user: “I’m looking for a restaurant”
- bot: “how about Chinese food?”
- user: “sure”
- bot: “here’s what I found ...”
Input data: just like dialogues in books:
Both user’s and bot’s responses are
defined by words they “say”
(not abstracted representations, e.g.,
intents )
Example from here
What is end-to-end training?
One model from text in context to next bot response text
Why end-to-end? Why not?
+ 1 model for all instead of 3
+ no need to annotate data with intents, slots, action labels and possible responses
+ data is more legible
- we need a lot of data to make end-to-end work properly
- we have less control over the outputs of the model
- story: end to end happy path
steps:
- user: “hi”
- bot: “hi!”
- user: “I’m looking for a restaurant”
- bot: “how about Chinese food?”
- user: “sure”
- bot: “here’s what I found ...”
Why now?
● On the one hand, fully unsupervised learning is seen as infeasible in production
● On the other hand, it is time that we start getting rid of intents in chatbots
Well… CHALLENGE ACCEPTED! 😋
- story: why we need end-to-end
steps:
- user: greet
- bot: utter_greet
- user: search_restaurant
- bot: utter_suggest_cuisine
- user: “I had that yesterday”
- bot: utter_suggest_cuisine
- story: why we need end-to-end
steps:
- user: greet
- bot: utter_greet
- user: search_restaurant
- bot: utter_suggest_cuisine
- user: “I had that yesterday but I am
always up for Thai”
- bot: utter_search_restaurant
Examples from here
What are main stages of research process at Rasa?
Research raw
idea🤓
If successful,
integrate the idea
into Rasa 👩‍🔧
Welcome the users to
use it as an experimental
feature 🎉
1st stage: Research raw idea🤓
Convince ourselves that end-to-end works overall
● Run experiments on multidomain text-to-text dataset in English -- MultiWOZ (Budzianowski et al,
2018)
● As Rasa retrieves the responses (rather than generating it), pick 20 random options
● Results looked good, so, we decided to proceed
Go through steps that our users would
● Took a bot built with a traditional pipeline and started “getting rid” of intents (Sara)
● Chatted with it as users would;
Model Hits@1 (20 candidates)
[Mehri et al., 2019]1
67.34
Ours 71.8
[Mehri et al., 2019]2
72.18
2nd stage: integrate the idea into Rasa 👩‍🔧
● Going fully end-to-end is a radical change of approach to how you keep your training data, which
models need to be trained etc.
⇒
How to combine these 2 approaches?
2nd stage: integrate the idea into Rasa 👩‍🔧
Example from here
- story: fully old format
steps:
- intent: greet
- action: utter_greet
- intent: search_restaurant
- action: utter_suggest_cuisine
- intent: affirm
- action: utter_search_restaurant
Usual “old” format
- story: fully old format
steps:
- intent: greet
- action: utter_greet
- intent: search_restaurant
- action: utter_suggest_cuisine
- user: “I am always up for
sushi”
- action: utter_search_restaurant
“Old” format with some user turns
in text (mix)
- story: end to end story
steps:
- user: “hi”
- bot: “hi!”
- user: “I’m looking for a
restaurant”
- bot: “how about Chinese food?”
- user: “sure”
- bot: “here’s what I found ...”
Fully end-to-end format
2nd stage: integrate the idea into Rasa 👩‍🔧
- story: fully old format
steps:
- intent: greet
- action: utter_greet
- intent: search_restaurant
- action: utter_suggest_cuisine
- user: “I am always up for
sushi”
Mixed dialogue
● we have NLU data mapping texts to intents;
● we have end-to-end stories, so, dialogue policy
will be able to take text into account in
prediction
1. NLU model predicts an intent
Example from here
How the prediction works
2. Dialogue policy can look directly at the text
3. Dialogue policy decides that the intent doesn’t
contain useful information and predicts next action
directly from text
search_restaurant
2nd stage: integrate the idea into Rasa 👩‍🔧
What does it mean?
● This makes intents optional
● We have a powerful dialogue policy which can be trained on intents and action names, texts of
phrases of the user and bot and any mix of those
After a couple of turns around the spiral of “this works but isn’t technically feasible” and “this is technically beautiful but
doesn’t work”...
Welcome the users to use it as an experimental
feature 🎉
Welcome! End-to-end is an experimental feature in Rasa 2.2
Thank you for your attention!
@erazumovskaia
evgeniar@yahoo.com
1 of 17

Recommended

UX Humor | Jokes and Funny Quotes by
UX Humor | Jokes and Funny QuotesUX Humor | Jokes and Funny Quotes
UX Humor | Jokes and Funny QuotesThink 360 Studio
112.1K views34 slides
Cómo adaptar Research UX en un entorno Agile by
Cómo adaptar Research UX en un entorno AgileCómo adaptar Research UX en un entorno Agile
Cómo adaptar Research UX en un entorno AgileRosa María Orellana Maldonado
717 views34 slides
A Workshop on UX Design and Storyboarding by
A Workshop on UX Design and StoryboardingA Workshop on UX Design and Storyboarding
A Workshop on UX Design and StoryboardingKevin Hoxie
3.5K views34 slides
UI vs UX workshop by
UI vs UX workshopUI vs UX workshop
UI vs UX workshopInova LLC
2.3K views16 slides
Intro to UX: Enterprise UX by
Intro to UX: Enterprise UXIntro to UX: Enterprise UX
Intro to UX: Enterprise UXBart Van Hecke
804 views184 slides
Simple Steps to UX/UI Web Design by
Simple Steps to UX/UI Web DesignSimple Steps to UX/UI Web Design
Simple Steps to UX/UI Web DesignKoombea
441.2K views31 slides

More Related Content

What's hot

Design process by
Design processDesign process
Design processLeandro Puca
941 views20 slides
Why User Experience Matters | By UX Professionals from Centerline Digital by
Why User Experience Matters | By UX Professionals from Centerline DigitalWhy User Experience Matters | By UX Professionals from Centerline Digital
Why User Experience Matters | By UX Professionals from Centerline DigitalCenterline Digital
13K views34 slides
Lean Product Design by
Lean Product DesignLean Product Design
Lean Product DesignLighting Beetle
3.6K views69 slides
Ux is not UI by
Ux is not UIUx is not UI
Ux is not UIBruno Mendes
19.6K views23 slides
Information Architecture Heuristics by
Information Architecture HeuristicsInformation Architecture Heuristics
Information Architecture HeuristicsAbby Covert
123.1K views35 slides
Understanding UI/UX Design by Aroyewun Babajide by
Understanding UI/UX Design by Aroyewun BabajideUnderstanding UI/UX Design by Aroyewun Babajide
Understanding UI/UX Design by Aroyewun BabajideBabajide Aroyewun
1.7K views27 slides

What's hot(20)

Why User Experience Matters | By UX Professionals from Centerline Digital by Centerline Digital
Why User Experience Matters | By UX Professionals from Centerline DigitalWhy User Experience Matters | By UX Professionals from Centerline Digital
Why User Experience Matters | By UX Professionals from Centerline Digital
Centerline Digital13K views
Information Architecture Heuristics by Abby Covert
Information Architecture HeuristicsInformation Architecture Heuristics
Information Architecture Heuristics
Abby Covert123.1K views
Understanding UI/UX Design by Aroyewun Babajide by Babajide Aroyewun
Understanding UI/UX Design by Aroyewun BabajideUnderstanding UI/UX Design by Aroyewun Babajide
Understanding UI/UX Design by Aroyewun Babajide
Babajide Aroyewun1.7K views
UX Design process, #UX, #Design Process, #Agile UX by Murali Krishna
UX Design process, #UX, #Design Process, #Agile UX UX Design process, #UX, #Design Process, #Agile UX
UX Design process, #UX, #Design Process, #Agile UX
Murali Krishna448 views
What is UI/UX and the Difference by Vibloo
What is UI/UX and the DifferenceWhat is UI/UX and the Difference
What is UI/UX and the Difference
Vibloo2.4K views
What is UX Design? by Baris Erkol
What is UX Design?What is UX Design?
What is UX Design?
Baris Erkol9.6K views
Designing user experience (ux) for digital products by Vijay Morampudi
Designing user experience (ux) for digital productsDesigning user experience (ux) for digital products
Designing user experience (ux) for digital products
Vijay Morampudi478 views
UX 101: A quick & dirty introduction to user experience strategy & design by Morgan McKeagney
UX 101: A quick & dirty introduction to user experience strategy & designUX 101: A quick & dirty introduction to user experience strategy & design
UX 101: A quick & dirty introduction to user experience strategy & design
Morgan McKeagney143.4K views
Product Design and UX / UI Design Process in Digital Product Development by Volodymyr Melnyk
Product Design and UX / UI Design Process in Digital Product DevelopmentProduct Design and UX / UI Design Process in Digital Product Development
Product Design and UX / UI Design Process in Digital Product Development
Volodymyr Melnyk960 views
LxD - Learner Experience Design by Julie Dirksen
LxD - Learner Experience DesignLxD - Learner Experience Design
LxD - Learner Experience Design
Julie Dirksen4.8K views
Evangelizing User Experience Design by Dustin DiTommaso
Evangelizing User Experience DesignEvangelizing User Experience Design
Evangelizing User Experience Design
Dustin DiTommaso5.1K views

Similar to End-to-end dialogue systems, or a feature which wasn’t meant to happen | Rasa Summit

Nlp workshop-share by
Nlp workshop-shareNlp workshop-share
Nlp workshop-shareQuantUniversity
1.3K views65 slides
Ed techjoker faculty presentation by
Ed techjoker faculty presentationEd techjoker faculty presentation
Ed techjoker faculty presentationbtopro
214 views114 slides
Free iOS apps for education by
Free iOS apps for educationFree iOS apps for education
Free iOS apps for educationAmy Terrell
134 views37 slides
Reactive declarative UI as code - DroidCon Vietnam 2019 by
Reactive declarative UI as code - DroidCon Vietnam 2019Reactive declarative UI as code - DroidCon Vietnam 2019
Reactive declarative UI as code - DroidCon Vietnam 2019oradoe
260 views69 slides
The story of a UI trainer by
The story of a UI trainerThe story of a UI trainer
The story of a UI trainerHarshal Patil
1K views70 slides
Module_4 English For IT Starter.pdf.pdf by
Module_4 English For IT Starter.pdf.pdfModule_4 English For IT Starter.pdf.pdf
Module_4 English For IT Starter.pdf.pdfAnna Gandrabura
94 views32 slides

Similar to End-to-end dialogue systems, or a feature which wasn’t meant to happen | Rasa Summit(20)

Ed techjoker faculty presentation by btopro
Ed techjoker faculty presentationEd techjoker faculty presentation
Ed techjoker faculty presentation
btopro214 views
Free iOS apps for education by Amy Terrell
Free iOS apps for educationFree iOS apps for education
Free iOS apps for education
Amy Terrell134 views
Reactive declarative UI as code - DroidCon Vietnam 2019 by oradoe
Reactive declarative UI as code - DroidCon Vietnam 2019Reactive declarative UI as code - DroidCon Vietnam 2019
Reactive declarative UI as code - DroidCon Vietnam 2019
oradoe260 views
Module_4 English For IT Starter.pdf.pdf by Anna Gandrabura
Module_4 English For IT Starter.pdf.pdfModule_4 English For IT Starter.pdf.pdf
Module_4 English For IT Starter.pdf.pdf
Anna Gandrabura94 views
Life skills for developers and architects by Rinka Singh
Life skills for developers and architectsLife skills for developers and architects
Life skills for developers and architects
Rinka Singh536 views
Building High-Quality Apps for Google Assistant by Peter Friese
Building High-Quality Apps for Google AssistantBuilding High-Quality Apps for Google Assistant
Building High-Quality Apps for Google Assistant
Peter Friese1.1K views
Session 1: UX Process + Interviewing by Leanna Gingras
Session 1: UX Process + InterviewingSession 1: UX Process + Interviewing
Session 1: UX Process + Interviewing
Leanna Gingras3.6K views
Impossible Programs by C4Media
Impossible ProgramsImpossible Programs
Impossible Programs
C4Media1.2K views
Voice usability testing with WOZ methodology - UX SCOT 2019 by Abi Reynolds
Voice usability testing with WOZ methodology - UX SCOT 2019Voice usability testing with WOZ methodology - UX SCOT 2019
Voice usability testing with WOZ methodology - UX SCOT 2019
Abi Reynolds144 views
[KDD 2018 tutorial] End to-end goal-oriented question answering systems by Qi He
[KDD 2018 tutorial] End to-end goal-oriented question answering systems[KDD 2018 tutorial] End to-end goal-oriented question answering systems
[KDD 2018 tutorial] End to-end goal-oriented question answering systems
Qi He8.1K views
Ux scot voice usability testing with woz - ar and sf - june 2019 by User Vision
Ux scot   voice usability testing with woz - ar and sf  - june 2019Ux scot   voice usability testing with woz - ar and sf  - june 2019
Ux scot voice usability testing with woz - ar and sf - june 2019
User Vision400 views
U Penn Wharton design challenge '17 by HJ Kwon
U Penn Wharton design challenge '17U Penn Wharton design challenge '17
U Penn Wharton design challenge '17
HJ Kwon883 views

More from Rasa Technologies

Six Steps to Conversation Driven Development by
Six Steps to Conversation Driven DevelopmentSix Steps to Conversation Driven Development
Six Steps to Conversation Driven DevelopmentRasa Technologies
355 views37 slides
Beyond Sentiment Analysis: Creating Engaging Conversational Experiences throu... by
Beyond Sentiment Analysis: Creating Engaging Conversational Experiences throu...Beyond Sentiment Analysis: Creating Engaging Conversational Experiences throu...
Beyond Sentiment Analysis: Creating Engaging Conversational Experiences throu...Rasa Technologies
250 views22 slides
Using Rasa to Power an Immersive Multimedia Conversational Experience | Rasa ... by
Using Rasa to Power an Immersive Multimedia Conversational Experience | Rasa ...Using Rasa to Power an Immersive Multimedia Conversational Experience | Rasa ...
Using Rasa to Power an Immersive Multimedia Conversational Experience | Rasa ...Rasa Technologies
178 views20 slides
How to Effectively Test Your Chatbot | Rasa Summit by
How to Effectively Test Your Chatbot  | Rasa SummitHow to Effectively Test Your Chatbot  | Rasa Summit
How to Effectively Test Your Chatbot | Rasa SummitRasa Technologies
142 views8 slides
Voice First: Ready Your Content to Serve 50% of Global Searches | Rasa Summit... by
Voice First: Ready Your Content to Serve 50% of Global Searches | Rasa Summit...Voice First: Ready Your Content to Serve 50% of Global Searches | Rasa Summit...
Voice First: Ready Your Content to Serve 50% of Global Searches | Rasa Summit...Rasa Technologies
112 views48 slides
The missing link: How AI can help create a safer society and better businesse... by
The missing link: How AI can help create a safer society and better businesse...The missing link: How AI can help create a safer society and better businesse...
The missing link: How AI can help create a safer society and better businesse...Rasa Technologies
120 views14 slides

More from Rasa Technologies(20)

Six Steps to Conversation Driven Development by Rasa Technologies
Six Steps to Conversation Driven DevelopmentSix Steps to Conversation Driven Development
Six Steps to Conversation Driven Development
Rasa Technologies355 views
Beyond Sentiment Analysis: Creating Engaging Conversational Experiences throu... by Rasa Technologies
Beyond Sentiment Analysis: Creating Engaging Conversational Experiences throu...Beyond Sentiment Analysis: Creating Engaging Conversational Experiences throu...
Beyond Sentiment Analysis: Creating Engaging Conversational Experiences throu...
Rasa Technologies250 views
Using Rasa to Power an Immersive Multimedia Conversational Experience | Rasa ... by Rasa Technologies
Using Rasa to Power an Immersive Multimedia Conversational Experience | Rasa ...Using Rasa to Power an Immersive Multimedia Conversational Experience | Rasa ...
Using Rasa to Power an Immersive Multimedia Conversational Experience | Rasa ...
Rasa Technologies178 views
How to Effectively Test Your Chatbot | Rasa Summit by Rasa Technologies
How to Effectively Test Your Chatbot  | Rasa SummitHow to Effectively Test Your Chatbot  | Rasa Summit
How to Effectively Test Your Chatbot | Rasa Summit
Rasa Technologies142 views
Voice First: Ready Your Content to Serve 50% of Global Searches | Rasa Summit... by Rasa Technologies
Voice First: Ready Your Content to Serve 50% of Global Searches | Rasa Summit...Voice First: Ready Your Content to Serve 50% of Global Searches | Rasa Summit...
Voice First: Ready Your Content to Serve 50% of Global Searches | Rasa Summit...
Rasa Technologies112 views
The missing link: How AI can help create a safer society and better businesse... by Rasa Technologies
The missing link: How AI can help create a safer society and better businesse...The missing link: How AI can help create a safer society and better businesse...
The missing link: How AI can help create a safer society and better businesse...
Rasa Technologies120 views
Boss - Bringing More Diversity to Tech | Rasa Summit by Rasa Technologies
Boss - Bringing More Diversity to Tech | Rasa SummitBoss - Bringing More Diversity to Tech | Rasa Summit
Boss - Bringing More Diversity to Tech | Rasa Summit
Rasa Technologies116 views
How Our Team Uses Rasa to Learn from Real Conversations | Rasa Summit by Rasa Technologies
How Our Team Uses Rasa to Learn from Real Conversations | Rasa SummitHow Our Team Uses Rasa to Learn from Real Conversations | Rasa Summit
How Our Team Uses Rasa to Learn from Real Conversations | Rasa Summit
Rasa Technologies136 views
Applying Conversational AI in the Enterprise by Rasa Technologies
Applying Conversational AI in the EnterpriseApplying Conversational AI in the Enterprise
Applying Conversational AI in the Enterprise
Rasa Technologies118 views
Supercharging User Interfaces with Rasa | Rasa Summit 2021 by Rasa Technologies
Supercharging User Interfaces with Rasa | Rasa Summit 2021Supercharging User Interfaces with Rasa | Rasa Summit 2021
Supercharging User Interfaces with Rasa | Rasa Summit 2021
STAR: A Schema-Guided Dialog Dataset for Transfer Learning | Rasa Summit 2021 by Rasa Technologies
STAR: A Schema-Guided Dialog Dataset for Transfer Learning | Rasa Summit 2021 STAR: A Schema-Guided Dialog Dataset for Transfer Learning | Rasa Summit 2021
STAR: A Schema-Guided Dialog Dataset for Transfer Learning | Rasa Summit 2021
Rasa Technologies215 views
Continuous Improvement of Conversational AI in Production | Rasa Summit by Rasa Technologies
Continuous Improvement of Conversational AI in Production | Rasa SummitContinuous Improvement of Conversational AI in Production | Rasa Summit
Continuous Improvement of Conversational AI in Production | Rasa Summit
Rasa Technologies130 views
Ethnobots: Reimagining Chatbots as Ethnographic Research Tools | Rasa Summit ... by Rasa Technologies
Ethnobots: Reimagining Chatbots as Ethnographic Research Tools | Rasa Summit ...Ethnobots: Reimagining Chatbots as Ethnographic Research Tools | Rasa Summit ...
Ethnobots: Reimagining Chatbots as Ethnographic Research Tools | Rasa Summit ...
The State of Conversation Design - Designing for the Conversational Future by Rasa Technologies
The State of Conversation Design - Designing for the Conversational FutureThe State of Conversation Design - Designing for the Conversational Future
The State of Conversation Design - Designing for the Conversational Future
Rasa Technologies353 views
Building an AI Assistant Factory - Rasa Summit 2021 by Rasa Technologies
Building an AI Assistant Factory - Rasa Summit 2021Building an AI Assistant Factory - Rasa Summit 2021
Building an AI Assistant Factory - Rasa Summit 2021
Rasa Technologies119 views
Building an End-to-End Test Automation Pipeline for Conversational AI | Rasa ... by Rasa Technologies
Building an End-to-End Test Automation Pipeline for Conversational AI | Rasa ...Building an End-to-End Test Automation Pipeline for Conversational AI | Rasa ...
Building an End-to-End Test Automation Pipeline for Conversational AI | Rasa ...
Rasa Technologies156 views
Deploy your Rasa Chatbots like a Boss with DevOps | Rasa Summit 2021 by Rasa Technologies
Deploy your Rasa Chatbots like a Boss with DevOps | Rasa Summit 2021Deploy your Rasa Chatbots like a Boss with DevOps | Rasa Summit 2021
Deploy your Rasa Chatbots like a Boss with DevOps | Rasa Summit 2021
Rasa Technologies184 views
What’s next in CDD: Intent Clashes and Selective Confidence | Rasa Summit 2021 by Rasa Technologies
What’s next in CDD: Intent Clashes and Selective Confidence | Rasa Summit 2021What’s next in CDD: Intent Clashes and Selective Confidence | Rasa Summit 2021
What’s next in CDD: Intent Clashes and Selective Confidence | Rasa Summit 2021
Rasa Technologies170 views

Recently uploaded

Setting Up Your First CloudStack Environment with Beginners Challenges - MD R... by
Setting Up Your First CloudStack Environment with Beginners Challenges - MD R...Setting Up Your First CloudStack Environment with Beginners Challenges - MD R...
Setting Up Your First CloudStack Environment with Beginners Challenges - MD R...ShapeBlue
105 views15 slides
Digital Personal Data Protection (DPDP) Practical Approach For CISOs by
Digital Personal Data Protection (DPDP) Practical Approach For CISOsDigital Personal Data Protection (DPDP) Practical Approach For CISOs
Digital Personal Data Protection (DPDP) Practical Approach For CISOsPriyanka Aash
103 views59 slides
NTGapps NTG LowCode Platform by
NTGapps NTG LowCode Platform NTGapps NTG LowCode Platform
NTGapps NTG LowCode Platform Mustafa Kuğu
287 views30 slides
KVM Security Groups Under the Hood - Wido den Hollander - Your.Online by
KVM Security Groups Under the Hood - Wido den Hollander - Your.OnlineKVM Security Groups Under the Hood - Wido den Hollander - Your.Online
KVM Security Groups Under the Hood - Wido den Hollander - Your.OnlineShapeBlue
154 views19 slides
Mitigating Common CloudStack Instance Deployment Failures - Jithin Raju - Sha... by
Mitigating Common CloudStack Instance Deployment Failures - Jithin Raju - Sha...Mitigating Common CloudStack Instance Deployment Failures - Jithin Raju - Sha...
Mitigating Common CloudStack Instance Deployment Failures - Jithin Raju - Sha...ShapeBlue
113 views18 slides
VNF Integration and Support in CloudStack - Wei Zhou - ShapeBlue by
VNF Integration and Support in CloudStack - Wei Zhou - ShapeBlueVNF Integration and Support in CloudStack - Wei Zhou - ShapeBlue
VNF Integration and Support in CloudStack - Wei Zhou - ShapeBlueShapeBlue
134 views54 slides

Recently uploaded(20)

Setting Up Your First CloudStack Environment with Beginners Challenges - MD R... by ShapeBlue
Setting Up Your First CloudStack Environment with Beginners Challenges - MD R...Setting Up Your First CloudStack Environment with Beginners Challenges - MD R...
Setting Up Your First CloudStack Environment with Beginners Challenges - MD R...
ShapeBlue105 views
Digital Personal Data Protection (DPDP) Practical Approach For CISOs by Priyanka Aash
Digital Personal Data Protection (DPDP) Practical Approach For CISOsDigital Personal Data Protection (DPDP) Practical Approach For CISOs
Digital Personal Data Protection (DPDP) Practical Approach For CISOs
Priyanka Aash103 views
NTGapps NTG LowCode Platform by Mustafa Kuğu
NTGapps NTG LowCode Platform NTGapps NTG LowCode Platform
NTGapps NTG LowCode Platform
Mustafa Kuğu287 views
KVM Security Groups Under the Hood - Wido den Hollander - Your.Online by ShapeBlue
KVM Security Groups Under the Hood - Wido den Hollander - Your.OnlineKVM Security Groups Under the Hood - Wido den Hollander - Your.Online
KVM Security Groups Under the Hood - Wido den Hollander - Your.Online
ShapeBlue154 views
Mitigating Common CloudStack Instance Deployment Failures - Jithin Raju - Sha... by ShapeBlue
Mitigating Common CloudStack Instance Deployment Failures - Jithin Raju - Sha...Mitigating Common CloudStack Instance Deployment Failures - Jithin Raju - Sha...
Mitigating Common CloudStack Instance Deployment Failures - Jithin Raju - Sha...
ShapeBlue113 views
VNF Integration and Support in CloudStack - Wei Zhou - ShapeBlue by ShapeBlue
VNF Integration and Support in CloudStack - Wei Zhou - ShapeBlueVNF Integration and Support in CloudStack - Wei Zhou - ShapeBlue
VNF Integration and Support in CloudStack - Wei Zhou - ShapeBlue
ShapeBlue134 views
2FA and OAuth2 in CloudStack - Andrija Panić - ShapeBlue by ShapeBlue
2FA and OAuth2 in CloudStack - Andrija Panić - ShapeBlue2FA and OAuth2 in CloudStack - Andrija Panić - ShapeBlue
2FA and OAuth2 in CloudStack - Andrija Panić - ShapeBlue
ShapeBlue75 views
Confidence in CloudStack - Aron Wagner, Nathan Gleason - Americ by ShapeBlue
Confidence in CloudStack - Aron Wagner, Nathan Gleason - AmericConfidence in CloudStack - Aron Wagner, Nathan Gleason - Americ
Confidence in CloudStack - Aron Wagner, Nathan Gleason - Americ
ShapeBlue58 views
Backup and Disaster Recovery with CloudStack and StorPool - Workshop - Venko ... by ShapeBlue
Backup and Disaster Recovery with CloudStack and StorPool - Workshop - Venko ...Backup and Disaster Recovery with CloudStack and StorPool - Workshop - Venko ...
Backup and Disaster Recovery with CloudStack and StorPool - Workshop - Venko ...
ShapeBlue114 views
DRBD Deep Dive - Philipp Reisner - LINBIT by ShapeBlue
DRBD Deep Dive - Philipp Reisner - LINBITDRBD Deep Dive - Philipp Reisner - LINBIT
DRBD Deep Dive - Philipp Reisner - LINBIT
ShapeBlue110 views
State of the Union - Rohit Yadav - Apache CloudStack by ShapeBlue
State of the Union - Rohit Yadav - Apache CloudStackState of the Union - Rohit Yadav - Apache CloudStack
State of the Union - Rohit Yadav - Apache CloudStack
ShapeBlue218 views
"Surviving highload with Node.js", Andrii Shumada by Fwdays
"Surviving highload with Node.js", Andrii Shumada "Surviving highload with Node.js", Andrii Shumada
"Surviving highload with Node.js", Andrii Shumada
Fwdays49 views
TrustArc Webinar - Managing Online Tracking Technology Vendors_ A Checklist f... by TrustArc
TrustArc Webinar - Managing Online Tracking Technology Vendors_ A Checklist f...TrustArc Webinar - Managing Online Tracking Technology Vendors_ A Checklist f...
TrustArc Webinar - Managing Online Tracking Technology Vendors_ A Checklist f...
TrustArc130 views
Migrating VMware Infra to KVM Using CloudStack - Nicolas Vazquez - ShapeBlue by ShapeBlue
Migrating VMware Infra to KVM Using CloudStack - Nicolas Vazquez - ShapeBlueMigrating VMware Infra to KVM Using CloudStack - Nicolas Vazquez - ShapeBlue
Migrating VMware Infra to KVM Using CloudStack - Nicolas Vazquez - ShapeBlue
ShapeBlue147 views
GDG Cloud Southlake 28 Brad Taylor and Shawn Augenstein Old Problems in the N... by James Anderson
GDG Cloud Southlake 28 Brad Taylor and Shawn Augenstein Old Problems in the N...GDG Cloud Southlake 28 Brad Taylor and Shawn Augenstein Old Problems in the N...
GDG Cloud Southlake 28 Brad Taylor and Shawn Augenstein Old Problems in the N...
James Anderson142 views
Zero to Cloud Hero: Crafting a Private Cloud from Scratch with XCP-ng, Xen Or... by ShapeBlue
Zero to Cloud Hero: Crafting a Private Cloud from Scratch with XCP-ng, Xen Or...Zero to Cloud Hero: Crafting a Private Cloud from Scratch with XCP-ng, Xen Or...
Zero to Cloud Hero: Crafting a Private Cloud from Scratch with XCP-ng, Xen Or...
ShapeBlue128 views
Webinar : Desperately Seeking Transformation - Part 2: Insights from leading... by The Digital Insurer
Webinar : Desperately Seeking Transformation - Part 2:  Insights from leading...Webinar : Desperately Seeking Transformation - Part 2:  Insights from leading...
Webinar : Desperately Seeking Transformation - Part 2: Insights from leading...
Updates on the LINSTOR Driver for CloudStack - Rene Peinthor - LINBIT by ShapeBlue
Updates on the LINSTOR Driver for CloudStack - Rene Peinthor - LINBITUpdates on the LINSTOR Driver for CloudStack - Rene Peinthor - LINBIT
Updates on the LINSTOR Driver for CloudStack - Rene Peinthor - LINBIT
ShapeBlue138 views
Import Export Virtual Machine for KVM Hypervisor - Ayush Pandey - University ... by ShapeBlue
Import Export Virtual Machine for KVM Hypervisor - Ayush Pandey - University ...Import Export Virtual Machine for KVM Hypervisor - Ayush Pandey - University ...
Import Export Virtual Machine for KVM Hypervisor - Ayush Pandey - University ...
ShapeBlue48 views

End-to-end dialogue systems, or a feature which wasn’t meant to happen | Rasa Summit

  • 1. End-to-end training, or a feature which wasn’t meant to happen Evgeniia (Genie) Razumovskaia PhD on Computation, Cognition and Language, University of Cambridge
  • 2. End-to-end training ● … as a new feature of Rasa ● … as an example of how research proceeds within Rasa
  • 3. What’s on: ● What is end-to-end training? ● Why end-to-end? Why now? ● What was the research process? ● What happens after research? (my biggest surprise 😅)
  • 4. What is end-to-end training? Traditional pipeline End-to-end training: predict the response text directly from the preceding text Image from here with little modifications
  • 5. What is end-to-end training? - story: end to end happy path steps: - user: “hi” - bot: “hi!” - user: “I’m looking for a restaurant” - bot: “how about Chinese food?” - user: “sure” - bot: “here’s what I found ...” Input data: just like dialogues in books: Both user’s and bot’s responses are defined by words they “say” (not abstracted representations, e.g., intents ) Example from here
  • 6. What is end-to-end training? One model from text in context to next bot response text
  • 7. Why end-to-end? Why not? + 1 model for all instead of 3 + no need to annotate data with intents, slots, action labels and possible responses + data is more legible - we need a lot of data to make end-to-end work properly - we have less control over the outputs of the model - story: end to end happy path steps: - user: “hi” - bot: “hi!” - user: “I’m looking for a restaurant” - bot: “how about Chinese food?” - user: “sure” - bot: “here’s what I found ...”
  • 8. Why now? ● On the one hand, fully unsupervised learning is seen as infeasible in production ● On the other hand, it is time that we start getting rid of intents in chatbots Well… CHALLENGE ACCEPTED! 😋 - story: why we need end-to-end steps: - user: greet - bot: utter_greet - user: search_restaurant - bot: utter_suggest_cuisine - user: “I had that yesterday” - bot: utter_suggest_cuisine - story: why we need end-to-end steps: - user: greet - bot: utter_greet - user: search_restaurant - bot: utter_suggest_cuisine - user: “I had that yesterday but I am always up for Thai” - bot: utter_search_restaurant Examples from here
  • 9. What are main stages of research process at Rasa? Research raw idea🤓 If successful, integrate the idea into Rasa 👩‍🔧 Welcome the users to use it as an experimental feature 🎉
  • 10. 1st stage: Research raw idea🤓 Convince ourselves that end-to-end works overall ● Run experiments on multidomain text-to-text dataset in English -- MultiWOZ (Budzianowski et al, 2018) ● As Rasa retrieves the responses (rather than generating it), pick 20 random options ● Results looked good, so, we decided to proceed Go through steps that our users would ● Took a bot built with a traditional pipeline and started “getting rid” of intents (Sara) ● Chatted with it as users would;
  • 11. Model Hits@1 (20 candidates) [Mehri et al., 2019]1 67.34 Ours 71.8 [Mehri et al., 2019]2 72.18
  • 12. 2nd stage: integrate the idea into Rasa 👩‍🔧 ● Going fully end-to-end is a radical change of approach to how you keep your training data, which models need to be trained etc. ⇒ How to combine these 2 approaches?
  • 13. 2nd stage: integrate the idea into Rasa 👩‍🔧 Example from here - story: fully old format steps: - intent: greet - action: utter_greet - intent: search_restaurant - action: utter_suggest_cuisine - intent: affirm - action: utter_search_restaurant Usual “old” format - story: fully old format steps: - intent: greet - action: utter_greet - intent: search_restaurant - action: utter_suggest_cuisine - user: “I am always up for sushi” - action: utter_search_restaurant “Old” format with some user turns in text (mix) - story: end to end story steps: - user: “hi” - bot: “hi!” - user: “I’m looking for a restaurant” - bot: “how about Chinese food?” - user: “sure” - bot: “here’s what I found ...” Fully end-to-end format
  • 14. 2nd stage: integrate the idea into Rasa 👩‍🔧 - story: fully old format steps: - intent: greet - action: utter_greet - intent: search_restaurant - action: utter_suggest_cuisine - user: “I am always up for sushi” Mixed dialogue ● we have NLU data mapping texts to intents; ● we have end-to-end stories, so, dialogue policy will be able to take text into account in prediction 1. NLU model predicts an intent Example from here How the prediction works 2. Dialogue policy can look directly at the text 3. Dialogue policy decides that the intent doesn’t contain useful information and predicts next action directly from text search_restaurant
  • 15. 2nd stage: integrate the idea into Rasa 👩‍🔧 What does it mean? ● This makes intents optional ● We have a powerful dialogue policy which can be trained on intents and action names, texts of phrases of the user and bot and any mix of those After a couple of turns around the spiral of “this works but isn’t technically feasible” and “this is technically beautiful but doesn’t work”...
  • 16. Welcome the users to use it as an experimental feature 🎉 Welcome! End-to-end is an experimental feature in Rasa 2.2
  • 17. Thank you for your attention! @erazumovskaia evgeniar@yahoo.com