A General Evaluation Framework to Assess Spoken Language Dialogue Systems: Experience with Call Center Agent Systems

Size: px
Start display at page:

Download "A General Evaluation Framework to Assess Spoken Language Dialogue Systems: Experience with Call Center Agent Systems"

Transcription

1 Conférence TALN 2000, Lausanne, octobre 2000 A General Evaluation Framework to Assess Spoken Language Dialogue Systems: Experience with Call Center Agent Systems Marcela Charfuelán, Cristina Esteban López Jose Relaño Gil, Ma. Carmen Rodríguez, Luis Hernández Gómez Dep. SSR ETSIT-UPM Ciudad Universitaria Madrid (Spain) marcela@gaps.ssr.upm.es Speech Tecnology Group, Telefónica Investigación y Desarrollo, S.A. C. Emilio Vargas, Madrid (Spain) Abstract In this paper we present our experience during the evaluation of two prototypes of call-center agents systems. We describe the general framework we have used during the collection and annotation of dialogue evaluation data bases. We also present our results using the well known PARADISE framework to derive a system performance function. The relative importance of different cost measurements for the SLDS s prototypes under evaluation is also discussed. 1. Introduction As Spoken Language Dialogue Systems (SLDSs) are becoming more and more attractive for a wide range of applications, there is an increasing demand on standardized benchmarks to test and compare their performance. The Spoken Language community has made significant progress towards this goal (Walker et al., 1998; Price et al., 1992; Minker, 1998), and most of the proposals for spoken dialogue evaluation are based on the use of information from properly designed evaluation dialogue corpora. Generally these corpora are extracted from log files as the evaluated system is working, and no specific nor standardized annotation procedures are used to represent the relevant information, though some proposals are presented in (DARPA, 1999; Isard et al., 1998; Dybkjoer et al., 1998). Our experience on SLDSs evaluation is related with the creation of call center agents based on a spoken dialogue system (Alvarez et al., 1996; Relaño et al., 1999) developed at Telefónica I+D of Spain: one called ATOS system whose domain is mainly telephonic functions, and the other called Voice PORTAL system whose domain is basically access to information through telephone. These agents were developed as prototypes for which we have designed an experimental evaluation procedure to organize the information collected in log files during an actual evaluation. This experimental evaluation framework has been already presented in

2 Marcela Charfuelán, Cristina Esteban López (Charfuelán et al., 2000), it could be summarized in three main aspects: an annotation scheme, an annotation tool and automatic extraction of dialogue metrics from annotated corpora. After we have collected our dialogue evaluation data bases we extract from them metrics and statistics like user and system turns average, number of tasks completed, user satisfaction etc. These metrics were used to calculate a predictive performance function of the system as it is proposed in PARADISE framework (Walker et al., 1998; Bonneau-Maynard et al., 2000). At First, in Section 2 the evaluation framework is briefly reviewed. In Section 3 we describe the dialogue system characteristics, architecture and functionality common to the two prototypes. Section 4 presents details of the evaluation environment and Section 5 of the dialogue evaluation databases obtained. Finally Section 6 contains an analysis of results after applying PARADISE framework and conclusions are made in Section 7 2. Overview of the Evaluation Framework 2.1. Annotation scheme As it is shown in Figure 1, we follow a two-step annotation process at utterance and dialogue level. At utterance level we perform two complementary tasks: processing logged information (for example: system response, recognizer s and parser s outputs) and including all manual objective or subjective information (such as the transcription of user utterances, or whether the recognizer s or parser s outputs correctly captured the task-related information in the utterance). The dialogue level is more global, here some information related to the dialogue structure is included (for example segments of dialogue corresponding to the starting and ending points of a particular SLDS task, or error recovery segments). After this dialogue structure mark-up is completed a set of simple automatic procedures are applied to obtain dialogue metrics and statistics as we will show afterwards in section 5. The output of the complete process is stored in annotated XML files. SLD System Audio file Log-file Annotation Levels Utterance Dialog Level Level User Questionnaires Annotator Annotated XML files Figure 1: Block diagram of the global annotation process 2.2. Annotation Methodology and Tools The annotation methodology combines manual procedures from a human annotator and automatic processing of annotated data. At utterance level, we needed an easy access to the audio speech file, then we developed an annotation tool for this level that we call ULAT (Utterance Level Annotation Tool). This tool let us:

3 L A TEX Style for TALN 2000 Manual transcription of user s turns having a controlled access to the audio file. Automatic extraction of information related to system turns, recognizer s and parser s outputs, and subjective information of the user from log files and external information files. Inclusion of subjective information from a human evaluator or annotator, for example, whether or not the user s concept (dialogue act) is lost after the speech recognizer and parser analysis. The inputs to the ULAT tool are: a recorded audio file of the dialogue, a log file provide by the system and an external information file (questionnaires). The annotator only has to mark a section of the speech wave, corresponding to a turn, listen to it and check if the information that the ULAT tool has presented for the turn is correct. The beginning and end samples are obtained by the tool and recorded automatically in the output file. At dialogue level we can use different XML tools because the files generated by the ULAT tool are in XML format. For example we have used the MATE Workbench (MATE, 1998) to annotate tasks (dialogue segments) an add the attributes of correctness, completion and user satisfaction of these segments. After that we used the LT XML (XML, 1999) tools and the developer s tool-kit (a C-based API) to extract metrics and statistics of evaluation from the XML data bases. 3. Dialogue System Characteristics The platform used to develop the call center agent prototypes ATOS system and Voice PORTAL system is based on the following main modules: Natural Language Speech Recognizer. Semantic Parser. Dialogue Management Module. Text-to-Speech System. The Natural Language Speech Recognition module has a vocabulary of 4500 words with 355 first names and 983 family names. It uses a language model based on trigrams working on parts-of-speech clustering. The Speech Recognition module uses context-dependent triphones represented through Hidden Markov Models. The Semantic Parser we use is a Spanish version of the PHOENIX parser developed at MIT, that can be described as a frame-based concept-spotting semantic parser. The Dialogue Manager is rule-based and its design is based on a collaborative dialogue model. According to the classification of Dialogue Systems proposed by J. Allen 1 it could be described as a system with topic-based performance capabilities, adaptive single task, a minimal pair clarification/correction dialogue manager and fixed mixed-initiative. 1 This classification was presented in the Tutorial: Dialogue Modelling by J. Allen, University of Rochester, in The ACL/EACL Workshop on Spoken Dialogue Systems. Madrid, Spain - July 11-12, 1997.

4 Marcela Charfuelán, Cristina Esteban López Finally the TTS system we use is the Spanish TTS developed by Telefónica Investigación y Desarrollo. This system is a diphone based system and includes rule-based prosodic modeling and LPC-based speech synthesis. 4. Evaluation Environment 4.1. Scenarios and Constrains The evaluation of the prototypes was done by selecting a set of different telephonic functions or tasks with different names or numbers as data constrains. The ATOS system was designed to execute twelve different tasks while the voice PORTAL could execute six. Table 1: Scenarios and Data Constraints in ATOS system Task Data T11 Phone Call Full name T12 Phone Call Phone number T13 Phone Call Extension number T41 Ask for information Electronic mail and a full name T42 Ask for information Office number and a full name T51 Multi-conference Two full names T52 Multi-conference Two phone numbers T53 Multi-conference Two extension numbers T61 Collect call Full name T62 Collect call Phone number T71 Change password Old and new password T81 Send a message Full name Table 2: Scenarios and Data Constraints in PORTAL system Task Data T1 Phone Call Full name T2 Where buy something Music or books T3 Ask for information Electronic mail or Office and a full name T4 News information National, International, Sports, Cultural, Weather and Society T5 Change password Old and new password T6 Send a message Full name All the full names are taken from a list or data base. The data in quotation marks (keywords) are information that is expected by the system. That not means that the user have to say them as if they were commands, the users are encouraged to use them in his own expressions in natural language. A population of 30 subjects was selected for the first field trial, who were novice users of the ATOS system, they executed 179 tasks (Approx. 3 hours 44 minutes of recording). For the second field trial with the Voice PORTAL system the population was of 17 novice users and they executed 50 tasks (Approx. 50 minutes 18 seconds of recording). Every subject involved

5 L A TEX Style for TALN 2000 in the evaluation processes was previously instructed in the basic functionality of the system and in the evaluation procedure. For every telephone call the dialogue system generated and stored the following information data: One speech audio file using 8 bits mu-law samples and a sampling frequency of 8 KHz. The whole dialogue was recorded in a single channel. An ASCII log file that included information of the dialogue system as it was working: the text of every system turn and the outputs of the recognizer and semantic parser for each user turn. Since each task in ATOS was simple, the testing procedure we used consisted of a sequence of six different functions executed by each subject. In this way, almost fifteen repetitions of each function or task were obtained. The same strategy was used for the Voice PORTAL system with the only difference that in the last each telephone call to the system was intended for more than one task User Satisfaction Measure This subjective metric was obtained through a survey made at the end of the tests. Each user was asked to complete a short questionnaire about the system. We made two questionnaires one for evaluation of each task and other for the global behaviour of the system. The questions for each task were: Could you complete the task? (Yes/No) How the system carried on this task? (1 to 10, 1 very bad and 10 excellent) The questions for the global evaluation were (1 very low or bad, 10 high or Excellent): Level of comprehension of the system prompts. Frequency in which you can t follow the dialogue. Level in which the system comprehend you. At which level the system become slow in its response time. Give us a score (1 to 10) of global evaluation of the system. Finally we averaged the answers for each task to obtain a user satisfaction score for each one. The question about the completion of the task was also useful to verify, during the annotation stage, if effectively the user completed the task (otherwise this metric could be not objective at all).

6 Marcela Charfuelán, Cristina Esteban López 4.3. Paradise Framework The PARADISE (PARAdigm for DIalogue System Evaluation) was proposed by (Walker et al., 1998) as a general framework for evaluating and comparing the performance of spoken dialogue agents. This framework uses methods from decision theory to combine a disparate set of performance measures into a single performance evaluation function. The objective of PAR- ADISE structure is to maximize user satisfaction through maximize task success and minimize costs (efficiency measures and qualitative measures). The performance equation is estimated using multivariate linear regression which provides different weights for each parameter in the performance equation. These weights give us an idea of the relative contribution of the success and cost factors to user satisfaction. The success at achieving the information requirements of the task is measured with the Kappa coefficient: Where is the proportion of times that one task have been successfully completed and is the proportion of times that one task is successful by chance. (We have given a slightly different meaning to and from the examples given in (Walker et al., 1998)). 5. Dialogue Evaluation Databases Following the two-stage annotation process at utterance and dialogue levels, we obtained two XML dialogue databases: EvalAtos and EvalPortal. Table 3 shows the performance metrics extracted from the EvalAtos data base which have been processed to get the mean values for each kind of task. The first column shows the scenarios and data constrains in each case. The mean dialogue metrics values per task for the Atos system were: 179 tasks executed (11.6 turns on average) from which 125 were completed or successfully completed (69.3%) and 54 not completed. The percentage of correct concepts (PCC) was 66.8%. We define a correct concept when the parser system could extract the name of the function or data from the recognized phrase for each user turn. As it will be discussed below it is important to notice that the PCC is not a precise reflect of the percentage of word recognition which in this case is 73.6%. Table 4 shows the Portal system performance metrics. The first column shows the scenarios in each case, as we have said here data constrains are a little bit different from previous data base, here they are more accurate and less variable than a full name or a number. The mean dialogue metrics values per task were: 50 tasks executed from which 44 were completed successfully and 6 not completed. The percentage of correct concepts here was 94.58% though the percentage of word recognition is 65.32%. 6. Analysis of Evaluation Results We have made a PARADISE paradigm analysis of the two systems described, based on the information extracted from the data bases. Our objectives were to estimate a performance function for user satisfaction (US) and compare the influence of different cost factors in both systems. For this we have calculated the kappa coefficient and as cost measures (or predictor factors) we have selected: average number of turns for each task (TT), percentage of correct concepts

7 L A TEX Style for TALN 2000 Table 3: Performance metrics for a Paradise case study, ATOS system : US = User Satisfaction, = Kappa coefficient, TT = Turns number for each Task, PCC = Percentage of Correct Concepts, PWR=Percentage of Word Recognition, TC = Percentage of task completed. Task US TT PCC PWR TC T11 Call (name) T12 Call (Phone number) T13 Call (extension) T42 Ask for information about Office number T51 Multi-conference (two names) T53 Multi-conference (two extensions) T41 Ask for information about electronic mail T61 Collect call (name) T62 Collect call (phone number) T71 Change password T81 Send a message T52 Multi-conference (two phone numbers) GLOBAL PERFORMANCE Table 4: Performance metrics for a Paradise case study, Voice PORTAL system : US = User Satisfaction, = Kappa coefficient, TT = Turns number for each Task, PCC = Percentage of Correct Concepts, PWR = Percentage of Word Recognition, TC = Percentage of task completed. Task US TT PCC PWR TC T1 Call (person name) T2 Buy music or books T3 Ask information about of a person T4 News information T5 Change user password T6 Send a message (person name) GLOBAL PERFORMANCE (PCC), percentage of word recognition (PWR), and percentage of complete tasks (TC). These are showed in tables 3 and 4. To make the regressions these values have to be normalized because the different factors are in different scales, then each factor x is normalized to its Z score: Where is the standard deviation for x. To calculate the regressions we can use several mathematical programs, these generally give additional information about the results, for example the standard error of prediction (p in our regressions) which gives an idea of how significant are the factors used. Another important information is the multiple correlation coefficient ( in our regressions) which gives an idea of the contribution of the combined factors to the variance of US. One of the techniques used in multiple linear regression to get the most predictive model

8 Marcela Charfuelán, Cristina Esteban López or equation is Forward Selection (Walpole & Myers, 1992). The procedure is basically to follow a sequence of regressions starting with individual factors. Each time the most significant factor (greater ) is selected and combined again with the others until the most significant combination is obtained. The estimations made for the ATOS system are the following: US = 0.572*TC *PCC (p= ) ( =81.07%) US = 0.608*TC *PCC *PWR (p= ) ( =82.63%) US = 0.421*TC *PCC * (p= ) ( =81.41%) US = 0.559*TC *PCC *TT (p= ) ( =93.37%) US = 0.542*TC *PCC *TT *PWR (p= ) ( =93.60%) Initially the factor most significant was TC that combined with PCC accounted for 81.07% of the variance in US, with a prediction error of p= When we combined these two factors with the others, PWR, and TT we observed the following: TT obtained the most significant contribution to the variance of US ( =93.37%), which is quite obvious because when the number of turns of one task become large and even if the dialogue system could recover from word recognition errors, the greater the number of turns is the less the US satisfaction could be. Respect to the others prediction factors, (PWR and ), they seem to be less significant but a little strange is the negative contribution of PWR, which could give us the erroneous idea that the less recognition the more user satisfaction. As we can see in the last regression involving all the factors, (except that is high-correlated with TC), the PWR factor contributes to the variance but with a very short positive weight respect to the others. That means that in this system it is more important PCC than PWR. In other words the US could be better with the same level of recognition if a robust parser is used. The estimations made for the Voice PORTAL system are the following: US = 1.013*PCC *PWR (p= ) ( =98.18%) US = 0.650*PCC *PWR * (p= ) ( =99.6%) US = 0.892*PCC *PWR *TT (p=0.0012) ( =98.82%) US = 0.650*PCC *PWR *TC (p= ) ( =99.6%) US = 0.624*PCC *PWR *TC *TT (p=0.0028) ( =99.80%) For the PORTAL system the most significant factor was PCC that combined with PWR accounted for 98.18% of the variance in US, with a prediction error of p= However the weight of PCC is greater than that of PWR, this could be explained because this system depends more on few specific keywords ( National, Sports, books, etc.) than on longer sequences of words like full names or telephone numbers. When we combined these two factors with the others, TT and TC we observed the following: TC and do not seem to contribute too much to the variance. Here again we could say that these two factors are correlated. But between PWR and TC we can observe that the weight of

9 L A TEX Style for TALN 2000 TC is greater than that of PWR, which confirm the fact that the semantic parser and the dialogue management module have done a very good work in spite of the level of word recognition. After all these regressions we find that one of the most important predictors of user satisfaction in both systems was the percentage of correct concepts (PCC), and one of the less important the percentage of word recognition (PWR). We could explain this because even though the PWR of the second system (65.3%) is less than the first (73.6%) its PCC is much better (94.5% compared with 66.8%) and it contributes more to user satisfaction. This also shows that it is more important the way we extract the concepts information (function and data) from the recognized phrase than the exact word recognition of the entire phrase. The task complexity is also important, for example in ATOS system a concept could be formed by a name a surname and even a number in the same utterance, then exact recognition of all the words is very important to infer the concept and data. On the other hand in voice PORTAL system only with the exact recognition of some keywords, that could be surrounded by some non-keywords, the system immediately could infer what kind of function is inquired. 7. Conclusion In this paper we have presented a general framework for dialogue annotation in the context of the evaluation of Spoken Language Agents. To examine the viability of the proposed coding scheme and annotation tools, they have been tested while evaluating two real prototypes of call-center agents using different simple dialogue metrics under the PARADISE framework. From the analysis of the experimental results we can say that the PARADISE methodology is useful to describe two apparently similar systems but with different behaviour in the field. Although more work it is necessary specially to determine which factors are correlated and in future trials verify the predictive performance of the equations obtained. Another noticeable conclusion of this work is the important role of the percentage of correct concepts over the percentage of word recognition. Then for future evaluations more emphasis will be necessary during the annotation stage of that we have called concepts (Antoine et al., 2000), because a more detailed classification and annotation could provide us with more information about the performance of the system. References ALVAREZ J., GIL J. C., CASAS C. C. & MERINO D. T. (1996). The natural language processing module for a voice asisted operator at telefónica i+d. In ICSLP 96, Philadelphia, USA. ANTOINE J.-Y., SIROUX J., CAELEN J., VILLANEAU J., GOULIAN J. & AHAFHAF M. (2000). Obtaining predictive results with and objective evaluation of spoken dialogue systems: experiments with the dcr assesment paradigm. In Proceedings of Second International Conference on Language Resources and Evaluation LREC-2000, Athens Greece. BONNEAU-MAYNARD H., DEVILLERS L. & ROSSET S. (2000). Predictive performance of dialog systems. In Proceedings of Second International Conference on Language Resources and Evaluation LREC-2000, Athens Greece. CHARFUELÁN M., RELAÑO GIL J., RODRÍGUEZ M. C., TAPIAS D. & GÓMEZ L. H. (2000). Dialogue annotation for language system evaluation. In Proceedings of Second International Conference on Language Resources and Evaluation LREC-2000, Athens Greece. DARPA (1999). Darpa communicator log standard version history.

10 Marcela Charfuelán, Cristina Esteban López DYBKJOER L., BERNSEN N. O., CARLSON R., CHASE L., DAHLBACK N., FAILENSCHMID K., HEID U., HEISTERKAMP P., JONSSON A., KAMP H., KARLSSON I., V. KUPPEVELT J., LAMEL L., PA- RAUBEK P. & WILLIAMS D. (1998). The disc approach to spoken language systems development and evaluation. In Proceedings of First International Conference on Language Resources and Evaluation, LREC-1998, Granada Spain. ISARD A., MCKELVIE D. & THOMPSON H. S. (1998). Towards a minimal standard for dialogue transcripts: A new sgml architecture for the hcrc map task corpus. In Proceeding of International Conference on Spoken Language Processing, Australia. MATE (1998). Mate. project overview. MINKER W. (1998). Evaluation methodologies for interactive speech systems. In Proceedings of First International Conference on Language Resources and Evaluation, LREC-1998, Granada Spain. RELAÑO GIL J., TAPIAS D., RODRÍGUEZ M. C., CHARFUELÁN M. & GÓMEZ L. H. (1999). Robust and flexible mixed-initiative dialogue for telephone services. In Ninth Conference of the European Chapter of the Association for Computational Linguistics, Bergen, Norway: Proceedings of EACL 99. PRICE P., HIRSCHMAN L., SHRIBERG E. & WADE E. (1992). Subject-based evaluation measures for interactive spoken language systems. In DARPA Proceedings of Speech and Natural Language Workshop. WALKER M., LITMAN D. J., KAMM C. A. & ABELLA A. (1998). Evaluating spoken dialogue agents with paradise: Two case studies. Computer Speech and Language, 12, WALPOLE R. E. & MYERS R. H. (1992). Probabilidad y Estadística. México: McGraw-Hill / Interamericana. XML L. (1999). Language technology group, lt xml version

1. Introduction to Spoken Dialogue Systems

1. Introduction to Spoken Dialogue Systems SoSe 2006 Projekt Sprachdialogsysteme 1. Introduction to Spoken Dialogue Systems Walther v. Hahn, Cristina Vertan {vhahn,vertan}@informatik.uni-hamburg.de Content What are Spoken dialogue systems? Types

More information

Language Technology II Language-Based Interaction Dialogue design, usability,evaluation. Word-error Rate. Basic Architecture of a Dialog System (3)

Language Technology II Language-Based Interaction Dialogue design, usability,evaluation. Word-error Rate. Basic Architecture of a Dialog System (3) Language Technology II Language-Based Interaction Dialogue design, usability,evaluation Manfred Pinkal Ivana Kruijff-Korbayová Course website: www.coli.uni-saarland.de/courses/late2 Basic Architecture

More information

Open-Source, Cross-Platform Java Tools Working Together on a Dialogue System

Open-Source, Cross-Platform Java Tools Working Together on a Dialogue System Open-Source, Cross-Platform Java Tools Working Together on a Dialogue System Oana NICOLAE Faculty of Mathematics and Computer Science, Department of Computer Science, University of Craiova, Romania oananicolae1981@yahoo.com

More information

How To Test A Speech Recognition System In Telef\U00F3Nica M\U00F3Viles Espa\U00F1A

How To Test A Speech Recognition System In Telef\U00F3Nica M\U00F3Viles Espa\U00F1A Methodology for rapid prototyping and testing of Speech Recognition user interfaces in Telefónica Móviles España Pedro Concejero, Juan José Rodríguez, (1) Daniel Tapias Merino (2) (1) pedro.concejero@tid.es

More information

Corpus Design for a Unit Selection Database

Corpus Design for a Unit Selection Database Corpus Design for a Unit Selection Database Norbert Braunschweiler Institute for Natural Language Processing (IMS) Stuttgart 8 th 9 th October 2002 BITS Workshop, München Norbert Braunschweiler Corpus

More information

VoiceXML Tutorial. Part 1: VoiceXML Basics and Simple Forms

VoiceXML Tutorial. Part 1: VoiceXML Basics and Simple Forms VoiceXML Tutorial Part 1: VoiceXML Basics and Simple Forms What is VoiceXML? XML Application W3C Standard Integration of Multiple Speech and Telephony Related Technologies Automated Speech Recognition

More information

VOICE INFORMATION RETRIEVAL FOR DOCUMENTS. Except where reference is made to the work of others, the work described in this thesis is.

VOICE INFORMATION RETRIEVAL FOR DOCUMENTS. Except where reference is made to the work of others, the work described in this thesis is. VOICE INFORMATION RETRIEVAL FOR DOCUMENTS Except where reference is made to the work of others, the work described in this thesis is my own or was done in collaboration with my advisory committee. Weihong

More information

D2.4: Two trained semantic decoders for the Appointment Scheduling task

D2.4: Two trained semantic decoders for the Appointment Scheduling task D2.4: Two trained semantic decoders for the Appointment Scheduling task James Henderson, François Mairesse, Lonneke van der Plas, Paola Merlo Distribution: Public CLASSiC Computational Learning in Adaptive

More information

ABSTRACT 2. SYSTEM OVERVIEW 1. INTRODUCTION. 2.1 Speech Recognition

ABSTRACT 2. SYSTEM OVERVIEW 1. INTRODUCTION. 2.1 Speech Recognition The CU Communicator: An Architecture for Dialogue Systems 1 Bryan Pellom, Wayne Ward, Sameer Pradhan Center for Spoken Language Research University of Colorado, Boulder Boulder, Colorado 80309-0594, USA

More information

Specialty Answering Service. All rights reserved.

Specialty Answering Service. All rights reserved. 0 Contents 1 Introduction... 2 1.1 Types of Dialog Systems... 2 2 Dialog Systems in Contact Centers... 4 2.1 Automated Call Centers... 4 3 History... 3 4 Designing Interactive Dialogs with Structured Data...

More information

PROMISE - A Procedure for Multimodal Interactive System Evaluation

PROMISE - A Procedure for Multimodal Interactive System Evaluation PROMISE - A Procedure for Multimodal Interactive System Evaluation Nicole Beringer Ute Kartal Katerina Louka Florian Schiel Uli Türk Ludwig-Maximilians-Universität München Report Nr. 23 Mai 2002 . Mai

More information

MIRACLE at VideoCLEF 2008: Classification of Multilingual Speech Transcripts

MIRACLE at VideoCLEF 2008: Classification of Multilingual Speech Transcripts MIRACLE at VideoCLEF 2008: Classification of Multilingual Speech Transcripts Julio Villena-Román 1,3, Sara Lana-Serrano 2,3 1 Universidad Carlos III de Madrid 2 Universidad Politécnica de Madrid 3 DAEDALUS

More information

An Arabic Text-To-Speech System Based on Artificial Neural Networks

An Arabic Text-To-Speech System Based on Artificial Neural Networks Journal of Computer Science 5 (3): 207-213, 2009 ISSN 1549-3636 2009 Science Publications An Arabic Text-To-Speech System Based on Artificial Neural Networks Ghadeer Al-Said and Moussa Abdallah Department

More information

2014/02/13 Sphinx Lunch

2014/02/13 Sphinx Lunch 2014/02/13 Sphinx Lunch Best Student Paper Award @ 2013 IEEE Workshop on Automatic Speech Recognition and Understanding Dec. 9-12, 2013 Unsupervised Induction and Filling of Semantic Slot for Spoken Dialogue

More information

Comparative Error Analysis of Dialog State Tracking

Comparative Error Analysis of Dialog State Tracking Comparative Error Analysis of Dialog State Tracking Ronnie W. Smith Department of Computer Science East Carolina University Greenville, North Carolina, 27834 rws@cs.ecu.edu Abstract A primary motivation

More information

Text-To-Speech Technologies for Mobile Telephony Services

Text-To-Speech Technologies for Mobile Telephony Services Text-To-Speech Technologies for Mobile Telephony Services Paulseph-John Farrugia Department of Computer Science and AI, University of Malta Abstract. Text-To-Speech (TTS) systems aim to transform arbitrary

More information

Information Leakage in Encrypted Network Traffic

Information Leakage in Encrypted Network Traffic Information Leakage in Encrypted Network Traffic Attacks and Countermeasures Scott Coull RedJack Joint work with: Charles Wright (MIT LL) Lucas Ballard (Google) Fabian Monrose (UNC) Gerald Masson (JHU)

More information

Develop Software that Speaks and Listens

Develop Software that Speaks and Listens Develop Software that Speaks and Listens Copyright 2011 Chant Inc. All rights reserved. Chant, SpeechKit, Getting the World Talking with Technology, talking man, and headset are trademarks or registered

More information

How To Recognize Voice Over Ip On Pc Or Mac Or Ip On A Pc Or Ip (Ip) On A Microsoft Computer Or Ip Computer On A Mac Or Mac (Ip Or Ip) On An Ip Computer Or Mac Computer On An Mp3

How To Recognize Voice Over Ip On Pc Or Mac Or Ip On A Pc Or Ip (Ip) On A Microsoft Computer Or Ip Computer On A Mac Or Mac (Ip Or Ip) On An Ip Computer Or Mac Computer On An Mp3 Recognizing Voice Over IP: A Robust Front-End for Speech Recognition on the World Wide Web. By C.Moreno, A. Antolin and F.Diaz-de-Maria. Summary By Maheshwar Jayaraman 1 1. Introduction Voice Over IP is

More information

A CHINESE SPEECH DATA WAREHOUSE

A CHINESE SPEECH DATA WAREHOUSE A CHINESE SPEECH DATA WAREHOUSE LUK Wing-Pong, Robert and CHENG Chung-Keng Department of Computing, Hong Kong Polytechnic University Tel: 2766 5143, FAX: 2774 0842, E-mail: {csrluk,cskcheng}@comp.polyu.edu.hk

More information

INF5820, Obligatory Assignment 3: Development of a Spoken Dialogue System

INF5820, Obligatory Assignment 3: Development of a Spoken Dialogue System INF5820, Obligatory Assignment 3: Development of a Spoken Dialogue System Pierre Lison October 29, 2014 In this project, you will develop a full, end-to-end spoken dialogue system for an application domain

More information

SOME ASPECTS OF ASR TRANSCRIPTION BASED UNSUPERVISED SPEAKER ADAPTATION FOR HMM SPEECH SYNTHESIS

SOME ASPECTS OF ASR TRANSCRIPTION BASED UNSUPERVISED SPEAKER ADAPTATION FOR HMM SPEECH SYNTHESIS SOME ASPECTS OF ASR TRANSCRIPTION BASED UNSUPERVISED SPEAKER ADAPTATION FOR HMM SPEECH SYNTHESIS Bálint Tóth, Tibor Fegyó, Géza Németh Department of Telecommunications and Media Informatics Budapest University

More information

31 Case Studies: Java Natural Language Tools Available on the Web

31 Case Studies: Java Natural Language Tools Available on the Web 31 Case Studies: Java Natural Language Tools Available on the Web Chapter Objectives Chapter Contents This chapter provides a number of sources for open source and free atural language understanding software

More information

Robust Methods for Automatic Transcription and Alignment of Speech Signals

Robust Methods for Automatic Transcription and Alignment of Speech Signals Robust Methods for Automatic Transcription and Alignment of Speech Signals Leif Grönqvist (lgr@msi.vxu.se) Course in Speech Recognition January 2. 2004 Contents Contents 1 1 Introduction 2 2 Background

More information

Robustness of a Spoken Dialogue Interface for a Personal Assistant

Robustness of a Spoken Dialogue Interface for a Personal Assistant Robustness of a Spoken Dialogue Interface for a Personal Assistant Anna Wong, Anh Nguyen and Wayne Wobcke School of Computer Science and Engineering University of New South Wales Sydney NSW 22, Australia

More information

VoiceXML-Based Dialogue Systems

VoiceXML-Based Dialogue Systems VoiceXML-Based Dialogue Systems Pavel Cenek Laboratory of Speech and Dialogue Faculty of Informatics Masaryk University Brno Agenda Dialogue system (DS) VoiceXML Frame-based DS in general 2 Computer based

More information

Word Completion and Prediction in Hebrew

Word Completion and Prediction in Hebrew Experiments with Language Models for בס"ד Word Completion and Prediction in Hebrew 1 Yaakov HaCohen-Kerner, Asaf Applebaum, Jacob Bitterman Department of Computer Science Jerusalem College of Technology

More information

LAGUARDIA COMMUNITY COLLEGE CITY UNIVERSITY OF NEW YORK DEPARTMENT OF MATHEMATICS, ENGINEERING, AND COMPUTER SCIENCE

LAGUARDIA COMMUNITY COLLEGE CITY UNIVERSITY OF NEW YORK DEPARTMENT OF MATHEMATICS, ENGINEERING, AND COMPUTER SCIENCE LAGUARDIA COMMUNITY COLLEGE CITY UNIVERSITY OF NEW YORK DEPARTMENT OF MATHEMATICS, ENGINEERING, AND COMPUTER SCIENCE MAT 119 STATISTICS AND ELEMENTARY ALGEBRA 5 Lecture Hours, 2 Lab Hours, 3 Credits Pre-

More information

Automatic Speech Recognition and Hybrid Machine Translation for High-Quality Closed-Captioning and Subtitling for Video Broadcast

Automatic Speech Recognition and Hybrid Machine Translation for High-Quality Closed-Captioning and Subtitling for Video Broadcast Automatic Speech Recognition and Hybrid Machine Translation for High-Quality Closed-Captioning and Subtitling for Video Broadcast Hassan Sawaf Science Applications International Corporation (SAIC) 7990

More information

The ROI. of Speech Tuning

The ROI. of Speech Tuning The ROI of Speech Tuning Executive Summary: Speech tuning is a process of improving speech applications after they have been deployed by reviewing how users interact with the system and testing changes.

More information

How to Get More Value from Your Survey Data

How to Get More Value from Your Survey Data Technical report How to Get More Value from Your Survey Data Discover four advanced analysis techniques that make survey research more effective Table of contents Introduction..............................................................2

More information

A System for Labeling Self-Repairs in Speech 1

A System for Labeling Self-Repairs in Speech 1 A System for Labeling Self-Repairs in Speech 1 John Bear, John Dowding, Elizabeth Shriberg, Patti Price 1. Introduction This document outlines a system for labeling self-repairs in spontaneous speech.

More information

Materials Software Systems Inc (MSSI). Enabling Speech on Touch Tone IVR White Paper

Materials Software Systems Inc (MSSI). Enabling Speech on Touch Tone IVR White Paper Materials Software Systems Inc (MSSI). Enabling Speech on Touch Tone IVR White Paper Reliable Customer Service and Automation is the key for Success in Hosted Interactive Voice Response Speech Enabled

More information

Dialog planning in VoiceXML

Dialog planning in VoiceXML Dialog planning in VoiceXML Csapó Tamás Gábor 4 January 2011 2. VoiceXML Programming Guide VoiceXML is an XML format programming language, describing the interactions between human

More information

2011 Springer-Verlag Berlin Heidelberg

2011 Springer-Verlag Berlin Heidelberg This document is published in: Novais, P. et al. (eds.) (2011). Ambient Intelligence - Software and Applications: 2nd International Symposium on Ambient Intelligence (ISAmI 2011). (Advances in Intelligent

More information

Standard Languages for Developing Multimodal Applications

Standard Languages for Developing Multimodal Applications Standard Languages for Developing Multimodal Applications James A. Larson Intel Corporation 16055 SW Walker Rd, #402, Beaverton, OR 97006 USA jim@larson-tech.com Abstract The World Wide Web Consortium

More information

Voice Driven Animation System

Voice Driven Animation System Voice Driven Animation System Zhijin Wang Department of Computer Science University of British Columbia Abstract The goal of this term project is to develop a voice driven animation system that could take

More information

Spot me if you can: Uncovering spoken phrases in encrypted VoIP conversations

Spot me if you can: Uncovering spoken phrases in encrypted VoIP conversations Spot me if you can: Uncovering spoken phrases in encrypted VoIP conversations C. Wright, L. Ballard, S. Coull, F. Monrose, G. Masson Talk held by Goran Doychev Selected Topics in Information Security and

More information

Open Source VoiceXML Interpreter over Asterisk for Use in IVR Applications

Open Source VoiceXML Interpreter over Asterisk for Use in IVR Applications Open Source VoiceXML Interpreter over Asterisk for Use in IVR Applications Lerato Lerato, Maletšabisa Molapo and Lehlohonolo Khoase Dept. of Maths and Computer Science, National University of Lesotho Roma

More information

Search and Information Retrieval

Search and Information Retrieval Search and Information Retrieval Search on the Web 1 is a daily activity for many people throughout the world Search and communication are most popular uses of the computer Applications involving search

More information

Emotion Detection from Speech

Emotion Detection from Speech Emotion Detection from Speech 1. Introduction Although emotion detection from speech is a relatively new field of research, it has many potential applications. In human-computer or human-human interaction

More information

AUTOMATIC PHONEME SEGMENTATION WITH RELAXED TEXTUAL CONSTRAINTS

AUTOMATIC PHONEME SEGMENTATION WITH RELAXED TEXTUAL CONSTRAINTS AUTOMATIC PHONEME SEGMENTATION WITH RELAXED TEXTUAL CONSTRAINTS PIERRE LANCHANTIN, ANDREW C. MORRIS, XAVIER RODET, CHRISTOPHE VEAUX Very high quality text-to-speech synthesis can be achieved by unit selection

More information

How do non-expert users exploit simultaneous inputs in multimodal interaction?

How do non-expert users exploit simultaneous inputs in multimodal interaction? How do non-expert users exploit simultaneous inputs in multimodal interaction? Knut Kvale, John Rugelbak and Ingunn Amdal 1 Telenor R&D, Norway knut.kvale@telenor.com, john.rugelbak@telenor.com, ingunn.amdal@tele.ntnu.no

More information

Bachelor's Degree in Business Administration and Master's Degree course description

Bachelor's Degree in Business Administration and Master's Degree course description Bachelor's Degree in Business Administration and Master's Degree course description Bachelor's Degree in Business Administration Department s Compulsory Requirements Course Description (402102) Principles

More information

IBM SPSS Direct Marketing 23

IBM SPSS Direct Marketing 23 IBM SPSS Direct Marketing 23 Note Before using this information and the product it supports, read the information in Notices on page 25. Product Information This edition applies to version 23, release

More information

IBM SPSS Direct Marketing 22

IBM SPSS Direct Marketing 22 IBM SPSS Direct Marketing 22 Note Before using this information and the product it supports, read the information in Notices on page 25. Product Information This edition applies to version 22, release

More information

Program curriculum for graduate studies in Speech and Music Communication

Program curriculum for graduate studies in Speech and Music Communication Program curriculum for graduate studies in Speech and Music Communication School of Computer Science and Communication, KTH (Translated version, November 2009) Common guidelines for graduate-level studies

More information

Membering T M : A Conference Call Service with Speaker-Independent Name Dialing on AIN

Membering T M : A Conference Call Service with Speaker-Independent Name Dialing on AIN PAGE 30 Membering T M : A Conference Call Service with Speaker-Independent Name Dialing on AIN Sung-Joon Park, Kyung-Ae Jang, Jae-In Kim, Myoung-Wan Koo, Chu-Shik Jhon Service Development Laboratory, KT,

More information

Flexible Spoken Dialogue System based on User Models and Dynamic Generation of VoiceXML Scripts

Flexible Spoken Dialogue System based on User Models and Dynamic Generation of VoiceXML Scripts Flexible Spoken Dialogue System based on User Models and Dynamic Generation of VoiceXML Scripts Kazunori Komatani Fumihiro Adachi Shinichi Ueno Tatsuya Kawahara Hiroshi G. Okuno Graduate School of Informatics

More information

SPEAKER IDENTITY INDEXING IN AUDIO-VISUAL DOCUMENTS

SPEAKER IDENTITY INDEXING IN AUDIO-VISUAL DOCUMENTS SPEAKER IDENTITY INDEXING IN AUDIO-VISUAL DOCUMENTS Mbarek Charhad, Daniel Moraru, Stéphane Ayache and Georges Quénot CLIPS-IMAG BP 53, 38041 Grenoble cedex 9, France Georges.Quenot@imag.fr ABSTRACT The

More information

Establishing the Uniqueness of the Human Voice for Security Applications

Establishing the Uniqueness of the Human Voice for Security Applications Proceedings of Student/Faculty Research Day, CSIS, Pace University, May 7th, 2004 Establishing the Uniqueness of the Human Voice for Security Applications Naresh P. Trilok, Sung-Hyuk Cha, and Charles C.

More information

CallAn: A Tool to Analyze Call Center Conversations

CallAn: A Tool to Analyze Call Center Conversations CallAn: A Tool to Analyze Call Center Conversations Balamurali AR, Frédéric Béchet And Benoit Favre Abstract Agent Quality Monitoring (QM) of customer calls is critical for call center companies. We present

More information

Collecting Polish German Parallel Corpora in the Internet

Collecting Polish German Parallel Corpora in the Internet Proceedings of the International Multiconference on ISSN 1896 7094 Computer Science and Information Technology, pp. 285 292 2007 PIPS Collecting Polish German Parallel Corpora in the Internet Monika Rosińska

More information

DATA ANALYSIS. QEM Network HBCU-UP Fundamentals of Education Research Workshop Gerunda B. Hughes, Ph.D. Howard University

DATA ANALYSIS. QEM Network HBCU-UP Fundamentals of Education Research Workshop Gerunda B. Hughes, Ph.D. Howard University DATA ANALYSIS QEM Network HBCU-UP Fundamentals of Education Research Workshop Gerunda B. Hughes, Ph.D. Howard University Quantitative Research What is Statistics? Statistics (as a subject) is the science

More information

Analysing Questionnaires using Minitab (for SPSS queries contact -) Graham.Currell@uwe.ac.uk

Analysing Questionnaires using Minitab (for SPSS queries contact -) Graham.Currell@uwe.ac.uk Analysing Questionnaires using Minitab (for SPSS queries contact -) Graham.Currell@uwe.ac.uk Structure As a starting point it is useful to consider a basic questionnaire as containing three main sections:

More information

AVoiceportal Enhanced by Semantic Processing and Affect Awareness

AVoiceportal Enhanced by Semantic Processing and Affect Awareness AVoiceportal Enhanced by Semantic Processing and Affect Awareness Felix Burkhardt, Joachim Stegmann, Markus VanBallegooy T-Systems International GmbH (felix.burkhardt joachim.stegmann markus.van-ballegooy)@t-systems.com

More information

Industrial Spoken Dialogue Design

Industrial Spoken Dialogue Design Enhanced Monitoring Tools and Online Dialogue Optimisation Merged into a New Spoken Dialogue System Design Experience Ghislain Putois Orange Labs Lannion, France Romain Laroche Orange Labs Issy-les-Moulineaux,

More information

Silvermine House Steenberg Office Park, Tokai 7945 Cape Town, South Africa Telephone: +27 21 702 4666 www.spss-sa.com

Silvermine House Steenberg Office Park, Tokai 7945 Cape Town, South Africa Telephone: +27 21 702 4666 www.spss-sa.com SPSS-SA Silvermine House Steenberg Office Park, Tokai 7945 Cape Town, South Africa Telephone: +27 21 702 4666 www.spss-sa.com SPSS-SA Training Brochure 2009 TABLE OF CONTENTS 1 SPSS TRAINING COURSES FOCUSING

More information

International Journal of Computer Trends and Technology (IJCTT) volume 4 Issue 8 August 2013

International Journal of Computer Trends and Technology (IJCTT) volume 4 Issue 8 August 2013 A Short-Term Traffic Prediction On A Distributed Network Using Multiple Regression Equation Ms.Sharmi.S 1 Research Scholar, MS University,Thirunelvelli Dr.M.Punithavalli Director, SREC,Coimbatore. Abstract:

More information

Speech Processing Applications in Quaero

Speech Processing Applications in Quaero Speech Processing Applications in Quaero Sebastian Stüker www.kit.edu 04.08 Introduction! Quaero is an innovative, French program addressing multimedia content! Speech technologies are part of the Quaero

More information

Turkish Radiology Dictation System

Turkish Radiology Dictation System Turkish Radiology Dictation System Ebru Arısoy, Levent M. Arslan Boaziçi University, Electrical and Electronic Engineering Department, 34342, Bebek, stanbul, Turkey arisoyeb@boun.edu.tr, arslanle@boun.edu.tr

More information

Correlational Research

Correlational Research Correlational Research Chapter Fifteen Correlational Research Chapter Fifteen Bring folder of readings The Nature of Correlational Research Correlational Research is also known as Associational Research.

More information

Carla Simões, t-carlas@microsoft.com. Speech Analysis and Transcription Software

Carla Simões, t-carlas@microsoft.com. Speech Analysis and Transcription Software Carla Simões, t-carlas@microsoft.com Speech Analysis and Transcription Software 1 Overview Methods for Speech Acoustic Analysis Why Speech Acoustic Analysis? Annotation Segmentation Alignment Speech Analysis

More information

OPTIMIZING YOUR MARKETING STRATEGY THROUGH MODELED TARGETING

OPTIMIZING YOUR MARKETING STRATEGY THROUGH MODELED TARGETING OPTIMIZING YOUR MARKETING STRATEGY THROUGH MODELED TARGETING 1 Introductions An insights-driven customer engagement firm Analytics-driven Marketing ROI focus Direct mail optimization 1.5 Billion 1:1 pieces

More information

Module Catalogue for the Bachelor Program in Computational Linguistics at the University of Heidelberg

Module Catalogue for the Bachelor Program in Computational Linguistics at the University of Heidelberg Module Catalogue for the Bachelor Program in Computational Linguistics at the University of Heidelberg March 1, 2007 The catalogue is organized into sections of (1) obligatory modules ( Basismodule ) that

More information

THE THIRD DIALOG STATE TRACKING CHALLENGE

THE THIRD DIALOG STATE TRACKING CHALLENGE THE THIRD DIALOG STATE TRACKING CHALLENGE Matthew Henderson 1, Blaise Thomson 2 and Jason D. Williams 3 1 Department of Engineering, University of Cambridge, UK 2 VocalIQ Ltd., Cambridge, UK 3 Microsoft

More information

An activity-based analysis of hands-on practice methods

An activity-based analysis of hands-on practice methods Journal of Computer Assisted Learning (2000) 16, 358-365 An activity-based analysis of hands-on practice methods S. Wiedenbeck, J.A. Zavala & J. Nawyn University of Nebraska Abstract The success of exploration-based

More information

GrammAds: Keyword and Ad Creative Generator for Online Advertising Campaigns

GrammAds: Keyword and Ad Creative Generator for Online Advertising Campaigns GrammAds: Keyword and Ad Creative Generator for Online Advertising Campaigns Stamatina Thomaidou 1,2, Konstantinos Leymonis 1,2, Michalis Vazirgiannis 1,2,3 Presented by: Fragkiskos Malliaros 2 1 : Athens

More information

PSG College of Technology, Coimbatore-641 004 Department of Computer & Information Sciences BSc (CT) G1 & G2 Sixth Semester PROJECT DETAILS.

PSG College of Technology, Coimbatore-641 004 Department of Computer & Information Sciences BSc (CT) G1 & G2 Sixth Semester PROJECT DETAILS. PSG College of Technology, Coimbatore-641 004 Department of Computer & Information Sciences BSc (CT) G1 & G2 Sixth Semester PROJECT DETAILS Project Project Title Area of Abstract No Specialization 1. Software

More information

Overview of iclef 2008: search log analysis for Multilingual Image Retrieval

Overview of iclef 2008: search log analysis for Multilingual Image Retrieval Overview of iclef 2008: search log analysis for Multilingual Image Retrieval Julio Gonzalo Paul Clough Jussi Karlgren UNED U. Sheffield SICS Spain United Kingdom Sweden julio@lsi.uned.es p.d.clough@sheffield.ac.uk

More information

interactive product brochure :: Nina: The Virtual Assistant for Mobile Customer Service Apps

interactive product brochure :: Nina: The Virtual Assistant for Mobile Customer Service Apps interactive product brochure :: Nina: The Virtual Assistant for Mobile Customer Service Apps This PDF contains embedded interactive features. Make sure to download and save the file to your computer to

More information

EFFECTIVENESS OF NOTE-TAKING SKILLS AND STUDENT'S CHARACTERISTICS ON LEARNING PERFORMANCE IN ONLINE COURSES

EFFECTIVENESS OF NOTE-TAKING SKILLS AND STUDENT'S CHARACTERISTICS ON LEARNING PERFORMANCE IN ONLINE COURSES EFFECTIVENESS OF NOTE-TAKING SKILLS AND STUDENT'S CHARACTERISTICS ON LEARNING PERFORMANCE IN ONLINE COURSES Nakayama, Minoru, Tokyo Institute of Technology, Ookayama 2-12-1 W9-107, Meguro, 152-8552 Tokyo,

More information

HOPS Project presentation

HOPS Project presentation HOPS Project presentation Enabling an Intelligent Natural Language Based Hub for the Deployment of Advanced Semantically Enriched Multi-channel Mass-scale Online Public Services IST-2002-507967 (HOPS)

More information

A Quagmire of Terminology: Verification & Validation, Testing, and Evaluation*

A Quagmire of Terminology: Verification & Validation, Testing, and Evaluation* From: FLAIRS-01 Proceedings. Copyright 2001, AAAI (www.aaai.org). All rights reserved. A Quagmire of Terminology: Verification & Validation, Testing, and Evaluation* Valerie Barr Department of Computer

More information

Evaluation of speech technologies

Evaluation of speech technologies CLARA Training course on evaluation of Human Language Technologies Evaluations and Language resources Distribution Agency November 27, 2012 Evaluation of speaker identification Speech technologies Outline

More information

Efficient diphone database creation for MBROLA, a multilingual speech synthesiser

Efficient diphone database creation for MBROLA, a multilingual speech synthesiser Efficient diphone database creation for, a multilingual speech synthesiser Institute of Linguistics Adam Mickiewicz University Poznań OWD 2010 Wisła-Kopydło, Poland Why? useful for testing speech models

More information

Micro blogs Oriented Word Segmentation System

Micro blogs Oriented Word Segmentation System Micro blogs Oriented Word Segmentation System Yijia Liu, Meishan Zhang, Wanxiang Che, Ting Liu, Yihe Deng Research Center for Social Computing and Information Retrieval Harbin Institute of Technology,

More information

Malay A. Dalal Madhav Erraguntla Perakath Benjamin. Knowledge Based Systems, Inc. (KBSI) College Station, TX 77840, U.S.A.

Malay A. Dalal Madhav Erraguntla Perakath Benjamin. Knowledge Based Systems, Inc. (KBSI) College Station, TX 77840, U.S.A. AN INTRODUCTION TO USING PROSIM FOR BUSINESS PROCESS SIMULATION AND ANALYSIS Malay A. Dalal Madhav Erraguntla Perakath Benjamin Knowledge Based Systems, Inc. (KBSI) College Station, TX 77840, U.S.A. ABSTRACT

More information

STATISTICS FOR PSYCHOLOGISTS

STATISTICS FOR PSYCHOLOGISTS STATISTICS FOR PSYCHOLOGISTS SECTION: STATISTICAL METHODS CHAPTER: REPORTING STATISTICS Abstract: This chapter describes basic rules for presenting statistical results in APA style. All rules come from

More information

Speech understanding in dialogue systems

Speech understanding in dialogue systems Speech understanding in dialogue systems Sergio Grau Puerto sgrau@dsic.upv.es Departament de Sistemes Informàtics i Computació Universitat Politècnica de València Sergio Grau Puerto. Carnegie Mellon: June

More information

Lean Six Sigma Analyze Phase Introduction. TECH 50800 QUALITY and PRODUCTIVITY in INDUSTRY and TECHNOLOGY

Lean Six Sigma Analyze Phase Introduction. TECH 50800 QUALITY and PRODUCTIVITY in INDUSTRY and TECHNOLOGY TECH 50800 QUALITY and PRODUCTIVITY in INDUSTRY and TECHNOLOGY Before we begin: Turn on the sound on your computer. There is audio to accompany this presentation. Audio will accompany most of the online

More information

customer care solutions

customer care solutions customer care solutions from Nuance white paper :: Understanding Natural Language Learning to speak customer-ese In recent years speech recognition systems have made impressive advances in their ability

More information

Architecture of an Ontology-Based Domain- Specific Natural Language Question Answering System

Architecture of an Ontology-Based Domain- Specific Natural Language Question Answering System Architecture of an Ontology-Based Domain- Specific Natural Language Question Answering System Athira P. M., Sreeja M. and P. C. Reghuraj Department of Computer Science and Engineering, Government Engineering

More information

Spoken Dialog Challenge 2010: Comparison of Live and Control Test Results

Spoken Dialog Challenge 2010: Comparison of Live and Control Test Results Spoken Dialog Challenge 2010: Comparison of Live and Control Test Results Alan W Black 1, Susanne Burger 1, Alistair Conkie 4, Helen Hastie 2, Simon Keizer 3, Oliver Lemon 2, Nicolas Merigaud 2, Gabriel

More information

Deposit Identification Utility and Visualization Tool

Deposit Identification Utility and Visualization Tool Deposit Identification Utility and Visualization Tool Colorado School of Mines Field Session Summer 2014 David Alexander Jeremy Kerr Luke McPherson Introduction Newmont Mining Corporation was founded in

More information

4/3/2014 STATISTICAL APPLICATIONS IN MARKET RESEARCH. Introductions. Tiffany Bonus, MS Chris Claeys, MS

4/3/2014 STATISTICAL APPLICATIONS IN MARKET RESEARCH. Introductions. Tiffany Bonus, MS Chris Claeys, MS STATISTICAL APPLICATIONS IN MARKET RESEARCH Introductions Tiffany Bonus, MS Chris Claeys, MS 1 Agenda What is Market Research? Job Responsibilities Important Skills Research Topics Statistical Applications

More information

Australian Standard. Interactive voice response systems user interface Speech recognition AS 5061 2008 AS 5061 2008

Australian Standard. Interactive voice response systems user interface Speech recognition AS 5061 2008 AS 5061 2008 AS 5061 2008 AS 5061 2008 Australian Standard Interactive voice response systems user interface Speech recognition This Australian Standard was prepared by Committee IT-022, Interactive Voice Response

More information

Abstract. Avaya Solution & Interoperability Test Lab

Abstract. Avaya Solution & Interoperability Test Lab Avaya Solution & Interoperability Test Lab Application Notes for LumenVox Automated Speech Recognizer, LumenVox Text-to-Speech Server and Call Progress Analysis with Avaya Aura Experience Portal Issue

More information

Data, Measurements, Features

Data, Measurements, Features Data, Measurements, Features Middle East Technical University Dep. of Computer Engineering 2009 compiled by V. Atalay What do you think of when someone says Data? We might abstract the idea that data are

More information

11. Analysis of Case-control Studies Logistic Regression

11. Analysis of Case-control Studies Logistic Regression Research methods II 113 11. Analysis of Case-control Studies Logistic Regression This chapter builds upon and further develops the concepts and strategies described in Ch.6 of Mother and Child Health:

More information

Master s Program in Information Systems

Master s Program in Information Systems The University of Jordan King Abdullah II School for Information Technology Department of Information Systems Master s Program in Information Systems 2006/2007 Study Plan Master Degree in Information Systems

More information

Transformation of Free-text Electronic Health Records for Efficient Information Retrieval and Support of Knowledge Discovery

Transformation of Free-text Electronic Health Records for Efficient Information Retrieval and Support of Knowledge Discovery Transformation of Free-text Electronic Health Records for Efficient Information Retrieval and Support of Knowledge Discovery Jan Paralic, Peter Smatana Technical University of Kosice, Slovakia Center for

More information

p. 3 p. 4 p. 93 p. 111 p. 119

p. 3 p. 4 p. 93 p. 111 p. 119 The relationship between minds-on and hands-on activity in instructional design : evidence from learning with interactive and non-interactive multimedia environments p. 3 Using automated capture in classrooms

More information

A secure face tracking system

A secure face tracking system International Journal of Information & Computation Technology. ISSN 0974-2239 Volume 4, Number 10 (2014), pp. 959-964 International Research Publications House http://www. irphouse.com A secure face tracking

More information

A Color Placement Support System for Visualization Designs Based on Subjective Color Balance

A Color Placement Support System for Visualization Designs Based on Subjective Color Balance A Color Placement Support System for Visualization Designs Based on Subjective Color Balance Eric Cooper and Katsuari Kamei College of Information Science and Engineering Ritsumeikan University Abstract:

More information

Support and Compatibility

Support and Compatibility Version 1.0 Frequently Asked Questions General What is Voiyager? Voiyager is a productivity platform for VoiceXML applications with Version 1.0 of Voiyager focusing on the complete development and testing

More information

Abstract. Key words: Voice-enabled Interfaces, Multimodal Interaction, Usability. 1. Introduction

Abstract. Key words: Voice-enabled Interfaces, Multimodal Interaction, Usability. 1. Introduction Evaluation of a multimodal Virtual Personal Assistant Glória Branco, Luís Almeida, Nuno Beires, Rui Gomes Voice ervices and Platforms - Portugal Telecom Inovação, Porto, Portugal [gloria, lalmeida, nbeires,

More information

Introduction to Engineering System Dynamics

Introduction to Engineering System Dynamics CHAPTER 0 Introduction to Engineering System Dynamics 0.1 INTRODUCTION The objective of an engineering analysis of a dynamic system is prediction of its behaviour or performance. Real dynamic systems are

More information

Winter 2016 Course Timetable. Legend: TIME: M = Monday T = Tuesday W = Wednesday R = Thursday F = Friday BREATH: M = Methodology: RA = Research Area

Winter 2016 Course Timetable. Legend: TIME: M = Monday T = Tuesday W = Wednesday R = Thursday F = Friday BREATH: M = Methodology: RA = Research Area Winter 2016 Course Timetable Legend: TIME: M = Monday T = Tuesday W = Wednesday R = Thursday F = Friday BREATH: M = Methodology: RA = Research Area Please note: Times listed in parentheses refer to the

More information

ISO and Industry Standards for User Centred Design

ISO and Industry Standards for User Centred Design ISO and Industry Standards for User Centred Design Nigel Bevan October 2000 www.usability.serco.com/trump nbevan@usability.serco.com Serco Usability Services, UK 2000 Serco Ltd. Reproduction permitted

More information