Dialogue Modeling 2. Indirect Requests “Can I have a cup of coffee?”  One approach to dealing with these kinds of requests is by plan-based inference.

Slides:



Advertisements
Similar presentations
C O N T E X T - F R E E LANGUAGES ( use a grammar to describe a language) 1.
Advertisements

1 CS101 Introduction to Computing Lecture 17 Algorithms II.
Timed Automata.
Speech Recognition Part 3 Back end processing. Speech recognition simplified block diagram Speech Capture Speech Capture Feature Extraction Feature Extraction.
© Rafael Moreno Esteban 2007 Can / Can’t Can is used to talk about ability and possibility, to ask for and give permission, and to make requests and offers.
 You will be able to: › Explain what is meant by an expert system and describe its components and applications.
Lecture 10 Conflict Resolution using a Problem-Solving approach Dr. Paul Wong D.Psyc.(Clinical) Centre for Suicide Research and Prevention.
Hidden Markov Models Ellen Walker Bioinformatics Hiram College, 2008.
Natural Language and Speech Processing Creation of computational models of the understanding and the generation of natural language. Different fields coming.
U1, Speech in the interface:2. Dialogue Management1 Module u1: Speech in the Interface 2: Dialogue Management Jacques Terken HG room 2:40 tel. (247) 5254.
1 Testing Oral Ability Pertemuan 22 Matakuliah: >/ > Tahun: >
NLP Pragmatics1 Pragmatics How do we use language to communicate? Dave Inman.
Making Making Arrangements By Ms. Terri Yueh. Making Arrangements Vocabulary & Expressions 1.Can we fix a meeting ? 2.Shall we arrange an appointment.
UTA Utah Transit Authority Kia Matthews. Introduction Utah Transit Authority is a provider of public transportation throughout the Wasatch Front of Utah,
Social Interactions: Inviting & Responding to Invitations
Reported Speech Roll No Presented By:- Class: Ixth “A”
Food. Are you hungry? I’m thinking of… A: I’m thinking of a food you eat in the morning. B: Cereal.
1 Making sound teacher judgments and moderating them Moderation for Primary Teachers Owhata School Staff meeting 26 September 2011.
Paper 2 Source Skills. Candidates’ weaknesses (according to examiners’ reports) Not supporting your answers with source detail Simply reproducing knowledge.
Future Time Simple Future: Will and be going to. Ali will finish his work tomorrow. Ali is going to finish his work tomorrow. Will and be going to express.
Comparative study of various Machine Learning methods For Telugu Part of Speech tagging -By Avinesh.PVS, Sudheer, Karthik IIIT - Hyderabad.
Welcome to my conference! February th grade Guadalupe.
Natural Language Processing Introduction. 2 Natural Language Processing We’re going to study what goes into getting computers to perform useful and interesting.
COURSE INTRODUCTION BUSINESS ENGLISH 2. Lecturer: BOGLARKA KISS KULENOVIĆ Office hours: Monday: 12:00 – 13:00 Tuesday: 10:00 – 12:00 Room: 20
Unit 2 A Flat World.  Objectives Objectives  FocusFocus  Warming up Warming up  7.1 Asking people to do things 7.1 Asking people to do things  7.2.
Dialogue Modeling. Task-Oriented Dialogues  Online appointment manager –Schedules and co-ordinates meetings between people User : Can you schedule a.
1 Instant Data Warehouse Utilities Extended (Again!!) 14/7/ Today I am pleased to announce the publishing of some fantastic new functionality for.
CS 6961: Structured Prediction Fall 2014 Course Information.
Business Communication 1. Guidelines 2 1. Use common courtesy in your request – ask rather than demand. 2. Include all the information the recipients.
COURSE INTRODUCTION BUSINESS ENGLISH 2. Lecturer: BOGLARKA KISS KULENOVIĆ Office hours: Tuesday: 10:00 – 12:00 1 more hour to be set for meetings Room:
Issues in Multiparty Dialogues Ronak Patel. Current Trend  Only two-party case (a person and a Dialog system  Multi party (more than two persons Ex.
Artificial Intelligence: Natural Language
COMPUTER ORGANIZATION AND ASSEMBLY LANGUAGE LECTURE BASED ON BOOLEAN EXPRESSIONS FOR COMBINATIONAL LOGIC SEPTEMBER 20, 2013.
Introduction to Dialogue Systems. User Input System Output ?
Agents that Reduce Work and Information Overload and Beyond Intelligent Interfaces Presented by Maulik Oza Department of Information and Computer Science.
Artificial Intelligence: Natural Language
For Monday Read chapter 26 Last Homework –Chapter 23, exercise 7.
ACE TESOL Diploma Program – London Language Institute OBJECTIVES You will understand: 1. The terminology and concepts of semantics, pragmatics and discourse.
Cover Letter YOUTH CENTRAL – Cover Letters & Templates
Software Development. Software Development Loop Design  Programmers need a solid foundation before they start coding anything  Understand the task.
Probabilistic reasoning over time Ch. 15, 17. Probabilistic reasoning over time So far, we’ve mostly dealt with episodic environments –Exceptions: games.
For Friday Finish chapter 23 Homework –Chapter 23, exercise 15.
1Computer Sciences Department. Book: INTRODUCTION TO THE THEORY OF COMPUTATION, SECOND EDITION, by: MICHAEL SIPSER Reference 3Computer Sciences Department.
© Rafael Moreno Esteban 2007 Can / Can’t Can is used to talk about ability(inability),rules®ulati ons, possibility, to ask for and give permission,
MODALS. PROBLEMS THAT STUDENTS HAVE USING MODALS.
CS 4705 Lecture 17 Semantic Analysis: Robust Semantics.
1 UNIT-3 KNOWLEDGE REPRESENTATION. 2 Agents that reason logically(Logical agents) A Knowledge based Agent The Wumpus world environment Representation,
For Monday Read chapter 26 Homework: –Chapter 23, exercises 8 and 9.
Finite Automata Great Theoretical Ideas In Computer Science Victor Adamchik Danny Sleator CS Spring 2010 Lecture 20Mar 30, 2010Carnegie Mellon.
© 2015 albert-learning.com How to talk to your boss How to talk to your boss!!
CHAPTER 9 ANNISA FAIZAH( ) RAHAJENG H. RARAS( ) ANA CLARISTI( ) DAMARINA( ) ASKING AND EXPLAINING.
The Simple Present: Yes / No Questions
LECTURE 5 Scanning. SYNTAX ANALYSIS We know from our previous lectures that the process of verifying the syntax of the program is performed in two stages:
Lecture 4 Page 1 CS 111 Online Modularity and Memory Clearly, programs must have access to memory We need abstractions that give them the required access.
Gluten Free Spaghetti alla Carbonara Recipe Hungry? With 5 ingredients (and I do mean 5 – not 5 plus salt, pepper and olive oil – but 5!) and about 15.
Job Applications LM10598 Designed by Learning Materials.
Lecture 4 Page 1 CS 111 Summer 2013 Scheduling CS 111 Operating Systems Peter Reiher.
What are they doing? sweep the floor clean the floor do the dishes.
Tasneem Ghnaimat. Language Model An abstract representation of a (natural) language. An approximation to real language Assume we have a set of sentences,
Unit 4 Why don’ t you talk to your parents? When you give advice, you can use the following expressions to be polite. Why don’t you do? Why not do …
Atlas.ti Training Manual Part 2: Codes. 2 PART 2: CODES What is a Code? A Code is an index category representing an.
Software Development.
Sequential Flexibility
PRAGMATICS 3.
CSE 105 theory of computation
Learning for Dialogue.
Natural Language Understanding
Pakitang-turo Dr. Lakangiting Garcia.
Answer these questions in full sentences in your books:
Information Retrieval
Presentation transcript:

Dialogue Modeling 2

Indirect Requests “Can I have a cup of coffee?”  One approach to dealing with these kinds of requests is by plan-based inference

Understanding an Indirect Request Asking if I am able to get a cup of coffee Interested in my ability to perform the action Being able to perform the action is a prerequisite for performing it Must want me to perform that action

Extends to other types of indirect requests too “My soup is cold.” People don’t like cold soup. I have cooked the meal, and Have access to the stove You want me to reheat your soup.

But requires lots of hand coded rules  E.g. : –All of the inferences rules we discussed yesterday  and for the previous example: –People don’t like cold soup –Knowledge that putting soup over an oven will heat it. –Knowing I am able to reheat soup

 Instead we can use surface level techniques : Cue-based Inference and Knowledge Poor Methods

Lecture Outline  Lexical cue-based Methods for Interpreting Dialogue Acts  Finite State Methods –Dialogue Managers –HMMs for Interpreting Dialogue Acts  Other Surface Level Dialogue Managers –Frame Based –Agenda Based

Cue-Based Methods for Interpretation  Instead of resolving fully, use surface features –“Can you”, “Please”  a Request –“Are you”  yes or no question –“Tell me”  Request

Context used also  S : “Would you like to make an appointment with John on Tuesday”  U : “Yes”  agreement  S : “Your appointment is with John on Tuesday.”  U : “Yes”  grounding

How would you build these models?  If you had a marked-up training set…  You could collect all of the sentences which you knew to be a particular dialogue act, then you could train a lexical mode : –D = argmax_d P(d | W) = argmax_d P(d) P (d | W)

And Context?  Use a finite-state framework

Lecture Outline  Lexical cue-based Methods for Interpreting Dialogue Acts  Finite State Methods –Dialogue Managers –HMMs for Interpreting Dialogue Acts  Other Surface Level Dialogue Managers –Frame Based –Agenda Based

Finite State Machines a b b a b a What is the language accepted by this machine? Equivalent to regular languages. Is this the minimal machine for this language?

Finite State Dialogue Model Do you want a pizza? yes no What size? large medium small … Bye! What toppings?

Pros  Simple to write  Very robust  Quick  System must direct entire conversation  User actions very limited Cons

Use Variables to shrink Automaton Do you want a pizza? What size? Size=response What toppings?

Lecture Outline  Lexical cue-based Methods for Interpreting Dialogue Acts  Finite State Methods –Dialogue Managers –HMMs for Interpreting Dialogue Acts  Other Surface Level Dialogue Managers –Frame Based –Agenda Based

HMMs  Hidden Markov Models –Probabilistic Finite State Automaton Are you hungry? Maybe :.8 Maybe :.2 Ok Would you like a bagel?

HMMs  So after one turn, don’t necessarily know what state you’re in Are you hungry? Maybe :.8 Maybe :.2 Ok Would you like a bagel?

Dialogue Model opening closing suggestconstrain accept reject

What State Are you In?  Combination of what words have been spoken : –P(E|D) : probability of words being spoken given that you’re in state D  And what a reasonable dialogue looks like –P(D) : probability of being in a dialogue state D

This model provides context  By knowing what state you were previously in, makes it easier to determine what the next state will be. –Different lexical frequencies in every state –Different probabilities for each sequences of dialogue states

Lecture Outline  Lexical cue-based Methods for Interpreting Dialogue Acts  Finite State Methods –Dialogue Managers –HMMs for Interpreting Dialogue Acts  Other Surface Level Dialogue Managers –Frame Based –Agenda Based

Frame-Based Dialogue Managers  For appointment scheduler, there are a few distinct things need to be decided  How would you write a finite state machine to get that information? People : ____ Time : _____ Location : _____

 This either : –Imposes a very rigid structure –Neglects useful shared structure  Want to flexibly ask for different pieces of the entire bit of information

Frame Driven Model  Solution : have a model where a frame drives the dialogue

Frame-Driven Dialogue User : I’d like to schedule an appointment. System : Who is the other party? User : John Calls up appointment frame People :____ Time : ____ Location : ___ People : John Time : __ Location : ___

Multiple Responses User : I’d like to schedule an appointment. System : Who is the other party? User : John, sometime on tuesday Calls up appointment frame People :____ Time : ____ Location : ___ People : John Time : Tuesday Location : ___

Grounding w/Frames –After each utterance,ground the request User : John, sometime on Tuesday System: Ok where do you want to meet John on Tuesday? –Alternatively can conform when form is finished System : Ok, I am scheduling you in Room 332 with John on Tuesday at 4pm