Joys and Pains of building a Dialog System

Slides:



Advertisements
Similar presentations
Querying with Units Getting what you asked for IVOA interoperability meeting M Hill (ROE), May 2004.
Advertisements

Non-Native Users in the Let s Go!! Spoken Dialogue System: Dealing with Linguistic Mismatch Antoine Raux & Maxine Eskenazi Language Technologies Institute.
Saying Hi. There are many ways to say “hi” to people.
Copyright © 2001 Bolton Institute Faculty of Technology Multimedia Integration and Applications Lecture 7: Prototype Review Damien Markey.
Learning in the Wild Satanjeev “Bano” Banerjee Dialogs on Dialog March 18 th, 2005 In the Meeting Room Scenario.
Making an E-learning Metacognition Activity 10/23 nick punt.
Dialogue and Information Retrieval Dialogs on Dialogs March all the way through April 2003.
Lecture 7: Prototype Review Damien Markey. Lecture 6: Prototype Review What makes a prototype successful Why a prototype is never a failure Review criteria.
5.3. S ECTION R OUNDUP Exploration of project hand-in and section roundup.
Goal Setting A projected state of affairs that a person plans to achieve A desired end point Desire, Intention, Purpose, Result It has been shown time.
Usability testing IS 403: User Interface Design Shaun Kane.
Adaptive Spoken Dialogue Systems & Computational Linguistics Diane J. Litman Dept. of Computer Science & Learning Research and Development Center University.
Academic Strategies Goal-Setting. Today we will discuss goal setting Select a goal that you want to accomplish Select a goal that you want to accomplish.
Grade Book Database Presentation Jeanne Winstead CINS 137.
Creating User Interfaces Ideas & Trends Homework: Post constructive comments. Work on project.
Sing, sing a song Sing out loud Sing out strong.
Debugging, bug finding and bug avoidance Part 1 Alan Dix
1 SYS366 Week 1 - Lecture 1 Introduction to Systems.
CSCI N201 Programming Concepts and Database 2 - STAIR Lingma Acheson Department of Computer and Information Science, IUPUI.
5.3. S ECTION R OUNDUP Exploration of project hand-in and section roundup.
Seminar on SDS Final class 5/4/2015. Topics discussed ASR, NLU in dialogue Dialogue management NLG, information presentation Evaluation Error recovery.
Steve Simon MVP SQL Server BI
Error Analysis Logic Errors.
Snapshots, checkpoints, rollback, and restart
DECISION-MAKING AND GOAL-SETTING
Mobile Testing - Bug Report
Reviewing Code A guide to smelling another developer’s source code.
Microsoft Access 2013 Bobby Wan.
1. BEGINNING STEPS QUESTION 3 – What in life is satisfying?
QUESTIONS ABOUT YOU SHOULD ASK.
Lesson 5-2 AP Computer Science Principles
Engineering and Debugging an App Chapter 15
Going Independent Pre discussion:
Keys to a Successful College Admission Essay
Steve Simon MVP SQL Server BI
Giraffes Can’t Dance.
Pre- production and planning
Evaluating Program Security
Introduction CSE 1310 – Introduction to Computers and Programming
Issues in Spoken Dialogue Systems
Spellodrome Willow Class Newsletter 4th April 2018
English Conversation I – Correction Techniques
Best Practices in Service Level Agreement
Strategies For Software Test Documentation
Top Tips Summary: Software Engineering in practice
CSCE 315 – Programming Studio, Fall 2017 Tanzir Ahmed
GRUNTMASTER6000 A leading innovation for future programmers.
Using the Six Traits of Writing
PostPC Computing Heuristic Evaluation Prof. Scott Kirkpatrick, HUJI
My Friendship Tools.
Using the Six Traits of Writing
Testing your digital portfolio
1.2 System Design Basics.
Error Analysis Runtime Errors.
S.T.A.I.R CSCI N207 Data Analysis Using Spreadsheet Lingma Acheson
All goals are not created equally.
Functions and their Combinations
Introduction to Access
How to Test Usability Modified from Keith Instone, Argus Associates.
Going Independent Pre discussion:
DO NOW 3/15/2016 Find the values
Using the Six Traits of Writing
Project Management.
Project Iterations.
Welcome! Let’s get creative!.
Paul Scerri and Nancy Reed
The STAR  Approach The interview questions tend to start with a variation of, "Tell me about a time when…" This may sound simple but, in the heat of the.
Break in the Cup.
Chapter 1: Creating a Program.
1 2 Do you know what the success criteria is? Yes No
Presentation transcript:

Joys and Pains of building a Dialog System Let’s Go!! Joys and Pains of building a Dialog System Antoine Dialogs on Dialogs Group April 18, 2003

Process Chart Task Specification Dialog Management Design Grammar Writing NLG Design Testing Evaluation Refinement TTS System Creation/Adaptation ASR Models Building

What is interesting? How do you make interaction as natural as possible? Design the initial task model Adding some disambiguation/confirmation agents Make the system sound smart (e.g. present results in a “natural” way, etc.)

What is interesting? (2) How do people actually dialogue with a machine? Use the system yourself Ask your friends to use it Design and perform experiments

What is time consuming? How do you get this f*!@?&g system to work? Install all the components Get your database up and running (~done) Get ASR to work properly (not there yet…) Get synthesis to work properly (idem) Debug pure bugs (typos in C, Perl, …)

Impressions Very rewarding: I can get real bus information from MY system!! Very frustrating: why can’t anyone else do???

Ideas, Advice Get a (very simple) FULLY running system as soon as possible (from ASR to Synthesis) Regularly fix milestones to have a (more and more complex) running system Try your system and have other people try it as much as possible Read the logs…

What about Let’s Go!! Main Problems Recognition is very poor for most users. Synthesis: Diphone voice not nice Limited domain: much better but fails (or very bad) when not in recorded data Backend: cannot interrupt queries (pb with misrecognitions that lead to super large queries e.g. “I’d like to go to Forbes”)

What about Let’s Go!! Next goal (for me): experiment to compare natives and non-natives Step 1: have the system work well enough with natives Step 2: get 10 natives and 20 non-natives use the system with scenarios Step 3: compare performance (success rate, number of misunderstandings…) Step 4: identify main sources of performance discrepancy Step 5: propose remedial measures