Errors in machine translations of English-Iraqi Arabic dialogues were analyzed at two different p... more Errors in machine translations of English-Iraqi Arabic dialogues were analyzed at two different points in the systems" development using HTER methods to identify errors and human annotations to refine TER annotations. Although the frequencies of errors in the more mature systems were lower, the proportions of error types exhibited little change. Results include high frequencies of pronoun errors in translations to English, high frequencies of subject person inflection in translations to Iraqi Arabic, similar frequencies of word order errors in both translation directions, and very low frequencies of polarity errors. The problems with many errors can be generalized as the need to insert lexemes not present in the source or vice versa, which includes errors in multi-word expressions. Discourse context will be required to resolve some problems with deictic elements like pronouns.
The communications of emergency response managers were tracked during simulated catastrophic even... more The communications of emergency response managers were tracked during simulated catastrophic events at a university campus in the Washington, D.C. region. Local, state, and federal response managers interacted with each other and with students using a variety of communication media in order to investigate the utility of new communication channels for emergency response management. Students and emergency managers interacted using a Twitter-like platform and a portal built with Ushahidi crowd-sourcing software. The emergency managers also used a chat interface that included private instant messaging, telephone, and the county's existing emergency web portal. Their media use was analyzed along with the functions of their communications, and the patterns that emerged are described and quantified.
Names of people, places, and organizations have unique linguistic properties, and they typically ... more Names of people, places, and organizations have unique linguistic properties, and they typically require special treatment in automatic processes. Appropriate processing of names is essential to achieve highquality information extraction, speech recognition, machine translation, and information management, yet most HLT applications provide limited specialized processing of names. Variation in the forms of names can make it difficult to retrieve names from data sources, to perform co-reference resolution across documents, or to associate instances of names with their representations in gazetteers and lexicons. Name matching has become critical in government contexts for checking watchlists and maintaining tax, health, and Social Security records. In commercial contexts, name matching is essential in credit, insurance, and legal applications.
The Defense Advanced Research Projects Agency (DARPA) Spoken Language Communication and Translati... more The Defense Advanced Research Projects Agency (DARPA) Spoken Language Communication and Translation System for Tactical Use (TRANSTAC) program (http://1.usa.gov/transtac) faced many challenges in applying automated measures of translation quality to Iraqi Arabic-English speech translation dialogues. Features of speech data in general and of Iraqi Arabic data in particular undermine basic assumptions of automated measures that depend on matching system outputs to reference translations. These features are described along with the challenges they present for evaluating machine translation quality using automated metrics. We show that scores for translation into Iraqi Arabic exhibit higher correlations with human judgments when they are computed from normalized system outputs and reference translations. Orthographic normalization, lexical normalization, and operations involving light stemming resulted in higher correlations with human judgments.
ACM Transactions on Asian Language Information Processing, 2011
Errors in machine translations of English-Iraqi Arabic dialogues were analyzed using the methods ... more Errors in machine translations of English-Iraqi Arabic dialogues were analyzed using the methods developed for the Human Translation Error Rate measure (HTER). Human annotations were used to refine the Translation Error Rate (TER) annotations. The analyses were performed on approximately 100 translations into each language from four translation systems. Results include high frequencies of pronoun errors and errors involving the
Abstract: The Spoken Language Communication and Translation System for Tactical Use (TRANSTAC) pr... more Abstract: The Spoken Language Communication and Translation System for Tactical Use (TRANSTAC) program is a Defense Advanced Research Agency (DARPA) program to create bidirectional speech-to-speech machine translation (MT) that will allow US Soldiers and Marines, speaking only English, to communicate, in tactical situations, with civilian populations who speak only other languages (for example, Iraqi Arabic). A key metric for the program is the odds of successfully transferring low-level concepts, defined as the source- ...
If a cyborg is any living entity in which the intervention of technology permanently alters the s... more If a cyborg is any living entity in which the intervention of technology permanently alters the structure and functioning of life processes, then most people are cyborgs in one way or another through eyeglasses and contact lenses, hearing aids, dental work, insulin, or birth control. Moreover, when people learn to read and write, they fundamentally alter the way their brains work by linking language processing mechanisms that evolved for auditory input to visual stimuli. Therefore, the technology of writing transforms the structure and functioning of our brains, just as it can transform human cultures and aspirations. This paper suggests that the technology of virtual writing provides new possibilities for communication that take advantage of processing strategies not available in ordinary speech.
This completes the conventions concerned solely with rendering audio tapes into written form. Con... more This completes the conventions concerned solely with rendering audio tapes into written form. Conventions for dividing speeches into utterance units are presented below. At the end of the section, some examples of the transcription format are presented.
who exhibit greater degrees of identi& exploration and role-taking skill participate in relations... more who exhibit greater degrees of identi& exploration and role-taking skill participate in relationships in which both individualiy and connectedness are expressed.
... types of interactions, the higher level functions that organize and orient the structure and ... more ... types of interactions, the higher level functions that organize and orient the structure and sequence of decisions are accomplished using different encoding ... we found between electronic and oral interactions, both types of discourse fit a generic decision-making schema that ...
Electronic Journal of Communication/La revue …, Jan 1, 1996
EJ550463 - Discourse Management Strategies in Face-to-Face and Computer-Mediated Decision Making ... more EJ550463 - Discourse Management Strategies in Face-to-Face and Computer-Mediated Decision Making Interactions.
Page 1. The discourse functions of OK SHERRI L CONDON1 Introduction Interjections like OK are ubi... more Page 1. The discourse functions of OK SHERRI L CONDON1 Introduction Interjections like OK are ubiquitous in ordinary conversation. Yet they are not easily classified into traditional taxonomies, such as parts of speech. They ...
... or closing some discourse structure, but in response to a first pairpart, the unmarked behavi... more ... or closing some discourse structure, but in response to a first pairpart, the unmarked behavior is the preferred (Levinson, 1983; Pomerantz, 1984) second ... of practices may have led to the evolution of ok as a transition marker from its use as an agreeing acknowledging complying ...
Page 1. Methodological Challenges of Selectivity in Family Interaction: Assessing Temporal Patter... more Page 1. Methodological Challenges of Selectivity in Family Interaction: Assessing Temporal Patterns of Individuation CATHERINE R. COOPER HAROLD D. GROTEVANT SHERRI M. CONDON University of Texas at Austin Twenty ...
The strategies adopted by dyads engaged in decision-making interactions are compared in face-to-f... more The strategies adopted by dyads engaged in decision-making interactions are compared in face-to-face, synchronous computer mediated, and asynchronous computer mediated (email) communication environments. The results show that the decisions required to ...
… 2001. Proceedings of the 34th Annual …, Jan 1, 2001
Turn-taking provides a basis for comparing interactions in different communication environments, ... more Turn-taking provides a basis for comparing interactions in different communication environments, and this paper demonstrates that readily observable features of turns can be linked to principles that organize and manage the interaction. Results are based on 150 decision-making interactions elicited in a face-to-face environment, an asynchronous, e-mail environment, and several types of synchronous computer-mediated environments. We show that three features of turns can be linked to discourse structure and function. First, as turns increase in size, participants switch from serial to parallel strategies to organize their decision-making. Second, pivot turns, which are turns that are much shorter than the turns that precede or follow them, can reflect the discourse functions of the relevant turns. Finally, turns can be used for measures of dominance based on turn size. We conclude that designers of communication systems can take advantage of peoples' ability to develop effective strategies for packaging messages in different environments. 0-7695-
This paper presents a solution to the problem of matching personal names in English to the same n... more This paper presents a solution to the problem of matching personal names in English to the same names represented in Arabic script. Standard string comparison measures perform poorly on this task due to varying transliteration conventions in both languages and the fact that Arabic script does not usually represent short vowels.
One of the most challenging tasks for uniformed service personnel serving in foreign countries is... more One of the most challenging tasks for uniformed service personnel serving in foreign countries is effective verbal communication with the local population. To remedy this problem, several companies and academic institutions have been funded to develop machine translation systems as part of the DARPA TRANSTAC (Spoken Language Communication and Translation System for Tactical Use) program. The goal of this program is to demonstrate capabilities to rapidly develop and field free-form, two-way translation systems that would enable speakers of different languages to communicate with one another in real-world tactical situations. DARPA has mandated that each TRANSTAC technology be evaluated numerous times throughout the life of the program and has tasked the National Institute of Standards and Technology (NIST) to lead this effort. This paper describes the experimental design methodology and test procedures from the most recent evaluation, conducted in July 2007, which focused on English to/from Iraqi Arabic.
Errors in machine translations of English-Iraqi Arabic dialogues were analyzed at two different p... more Errors in machine translations of English-Iraqi Arabic dialogues were analyzed at two different points in the systems" development using HTER methods to identify errors and human annotations to refine TER annotations. Although the frequencies of errors in the more mature systems were lower, the proportions of error types exhibited little change. Results include high frequencies of pronoun errors in translations to English, high frequencies of subject person inflection in translations to Iraqi Arabic, similar frequencies of word order errors in both translation directions, and very low frequencies of polarity errors. The problems with many errors can be generalized as the need to insert lexemes not present in the source or vice versa, which includes errors in multi-word expressions. Discourse context will be required to resolve some problems with deictic elements like pronouns.
The communications of emergency response managers were tracked during simulated catastrophic even... more The communications of emergency response managers were tracked during simulated catastrophic events at a university campus in the Washington, D.C. region. Local, state, and federal response managers interacted with each other and with students using a variety of communication media in order to investigate the utility of new communication channels for emergency response management. Students and emergency managers interacted using a Twitter-like platform and a portal built with Ushahidi crowd-sourcing software. The emergency managers also used a chat interface that included private instant messaging, telephone, and the county's existing emergency web portal. Their media use was analyzed along with the functions of their communications, and the patterns that emerged are described and quantified.
Names of people, places, and organizations have unique linguistic properties, and they typically ... more Names of people, places, and organizations have unique linguistic properties, and they typically require special treatment in automatic processes. Appropriate processing of names is essential to achieve highquality information extraction, speech recognition, machine translation, and information management, yet most HLT applications provide limited specialized processing of names. Variation in the forms of names can make it difficult to retrieve names from data sources, to perform co-reference resolution across documents, or to associate instances of names with their representations in gazetteers and lexicons. Name matching has become critical in government contexts for checking watchlists and maintaining tax, health, and Social Security records. In commercial contexts, name matching is essential in credit, insurance, and legal applications.
The Defense Advanced Research Projects Agency (DARPA) Spoken Language Communication and Translati... more The Defense Advanced Research Projects Agency (DARPA) Spoken Language Communication and Translation System for Tactical Use (TRANSTAC) program (http://1.usa.gov/transtac) faced many challenges in applying automated measures of translation quality to Iraqi Arabic-English speech translation dialogues. Features of speech data in general and of Iraqi Arabic data in particular undermine basic assumptions of automated measures that depend on matching system outputs to reference translations. These features are described along with the challenges they present for evaluating machine translation quality using automated metrics. We show that scores for translation into Iraqi Arabic exhibit higher correlations with human judgments when they are computed from normalized system outputs and reference translations. Orthographic normalization, lexical normalization, and operations involving light stemming resulted in higher correlations with human judgments.
ACM Transactions on Asian Language Information Processing, 2011
Errors in machine translations of English-Iraqi Arabic dialogues were analyzed using the methods ... more Errors in machine translations of English-Iraqi Arabic dialogues were analyzed using the methods developed for the Human Translation Error Rate measure (HTER). Human annotations were used to refine the Translation Error Rate (TER) annotations. The analyses were performed on approximately 100 translations into each language from four translation systems. Results include high frequencies of pronoun errors and errors involving the
Abstract: The Spoken Language Communication and Translation System for Tactical Use (TRANSTAC) pr... more Abstract: The Spoken Language Communication and Translation System for Tactical Use (TRANSTAC) program is a Defense Advanced Research Agency (DARPA) program to create bidirectional speech-to-speech machine translation (MT) that will allow US Soldiers and Marines, speaking only English, to communicate, in tactical situations, with civilian populations who speak only other languages (for example, Iraqi Arabic). A key metric for the program is the odds of successfully transferring low-level concepts, defined as the source- ...
If a cyborg is any living entity in which the intervention of technology permanently alters the s... more If a cyborg is any living entity in which the intervention of technology permanently alters the structure and functioning of life processes, then most people are cyborgs in one way or another through eyeglasses and contact lenses, hearing aids, dental work, insulin, or birth control. Moreover, when people learn to read and write, they fundamentally alter the way their brains work by linking language processing mechanisms that evolved for auditory input to visual stimuli. Therefore, the technology of writing transforms the structure and functioning of our brains, just as it can transform human cultures and aspirations. This paper suggests that the technology of virtual writing provides new possibilities for communication that take advantage of processing strategies not available in ordinary speech.
This completes the conventions concerned solely with rendering audio tapes into written form. Con... more This completes the conventions concerned solely with rendering audio tapes into written form. Conventions for dividing speeches into utterance units are presented below. At the end of the section, some examples of the transcription format are presented.
who exhibit greater degrees of identi& exploration and role-taking skill participate in relations... more who exhibit greater degrees of identi& exploration and role-taking skill participate in relationships in which both individualiy and connectedness are expressed.
... types of interactions, the higher level functions that organize and orient the structure and ... more ... types of interactions, the higher level functions that organize and orient the structure and sequence of decisions are accomplished using different encoding ... we found between electronic and oral interactions, both types of discourse fit a generic decision-making schema that ...
Electronic Journal of Communication/La revue …, Jan 1, 1996
EJ550463 - Discourse Management Strategies in Face-to-Face and Computer-Mediated Decision Making ... more EJ550463 - Discourse Management Strategies in Face-to-Face and Computer-Mediated Decision Making Interactions.
Page 1. The discourse functions of OK SHERRI L CONDON1 Introduction Interjections like OK are ubi... more Page 1. The discourse functions of OK SHERRI L CONDON1 Introduction Interjections like OK are ubiquitous in ordinary conversation. Yet they are not easily classified into traditional taxonomies, such as parts of speech. They ...
... or closing some discourse structure, but in response to a first pairpart, the unmarked behavi... more ... or closing some discourse structure, but in response to a first pairpart, the unmarked behavior is the preferred (Levinson, 1983; Pomerantz, 1984) second ... of practices may have led to the evolution of ok as a transition marker from its use as an agreeing acknowledging complying ...
Page 1. Methodological Challenges of Selectivity in Family Interaction: Assessing Temporal Patter... more Page 1. Methodological Challenges of Selectivity in Family Interaction: Assessing Temporal Patterns of Individuation CATHERINE R. COOPER HAROLD D. GROTEVANT SHERRI M. CONDON University of Texas at Austin Twenty ...
The strategies adopted by dyads engaged in decision-making interactions are compared in face-to-f... more The strategies adopted by dyads engaged in decision-making interactions are compared in face-to-face, synchronous computer mediated, and asynchronous computer mediated (email) communication environments. The results show that the decisions required to ...
… 2001. Proceedings of the 34th Annual …, Jan 1, 2001
Turn-taking provides a basis for comparing interactions in different communication environments, ... more Turn-taking provides a basis for comparing interactions in different communication environments, and this paper demonstrates that readily observable features of turns can be linked to principles that organize and manage the interaction. Results are based on 150 decision-making interactions elicited in a face-to-face environment, an asynchronous, e-mail environment, and several types of synchronous computer-mediated environments. We show that three features of turns can be linked to discourse structure and function. First, as turns increase in size, participants switch from serial to parallel strategies to organize their decision-making. Second, pivot turns, which are turns that are much shorter than the turns that precede or follow them, can reflect the discourse functions of the relevant turns. Finally, turns can be used for measures of dominance based on turn size. We conclude that designers of communication systems can take advantage of peoples' ability to develop effective strategies for packaging messages in different environments. 0-7695-
This paper presents a solution to the problem of matching personal names in English to the same n... more This paper presents a solution to the problem of matching personal names in English to the same names represented in Arabic script. Standard string comparison measures perform poorly on this task due to varying transliteration conventions in both languages and the fact that Arabic script does not usually represent short vowels.
One of the most challenging tasks for uniformed service personnel serving in foreign countries is... more One of the most challenging tasks for uniformed service personnel serving in foreign countries is effective verbal communication with the local population. To remedy this problem, several companies and academic institutions have been funded to develop machine translation systems as part of the DARPA TRANSTAC (Spoken Language Communication and Translation System for Tactical Use) program. The goal of this program is to demonstrate capabilities to rapidly develop and field free-form, two-way translation systems that would enable speakers of different languages to communicate with one another in real-world tactical situations. DARPA has mandated that each TRANSTAC technology be evaluated numerous times throughout the life of the program and has tasked the National Institute of Standards and Technology (NIST) to lead this effort. This paper describes the experimental design methodology and test procedures from the most recent evaluation, conducted in July 2007, which focused on English to/from Iraqi Arabic.
Uploads
Papers by Sherri Condon