Hanady Ahmed | Qatar University (original) (raw)
Papers by Hanady Ahmed
International Journal of Speech Technology
Bookmarks Related papers MentionsView impact
Proceedings of The 12th International Workshop on Semantic Evaluation
Bookmarks Related papers MentionsView impact
Information, Communication & Society
Bookmarks Related papers MentionsView impact
Proceedings of the Third Arabic Natural Language Processing Workshop
Bookmarks Related papers MentionsView impact
Bookmarks Related papers MentionsView impact
Information
Assigning sentiment labels to documents is, at first sight, a standard multi-label classification... more Assigning sentiment labels to documents is, at first sight, a standard multi-label classification task. Many approaches have been used for this task, but the current state-of-the-art solutions use deep neural networks (DNNs). As such, it seems likely that standard machine learning algorithms, such as these, will provide an effective approach. We describe an alternative approach, involving the use of probabilities to construct a weighted lexicon of sentiment terms, then modifying the lexicon and calculating optimal thresholds for each class. We show that this approach outperforms the use of DNNs and other standard algorithms. We believe that DNNs are not a universal panacea and that paying attention to the nature of the data that you are trying to learn from can be more important than trying out ever more powerful general purpose machine learning algorithms.
Bookmarks Related papers MentionsView impact
Computer Speech & Language, 2008
... b University of Alexandria, Egypt. ... Which of these gets chosen depends on the specific ten... more ... b University of Alexandria, Egypt. ... Which of these gets chosen depends on the specific tense markers that get added and on whether the verb is used in an active or apassive sentence. ... Other affixes cannot be settled so early. ...
Bookmarks Related papers MentionsView impact
Résumé - Abstract The work,described here is part of an attempt to provide a text-to-speech (TTS)... more Résumé - Abstract The work,described here is part of an attempt to provide a text-to-speech (TTS) system for
Bookmarks Related papers MentionsView impact
We have described elsewhere a text-to-speech (TTS) system for Modern Standard Arabic which impose... more We have described elsewhere a text-to-speech (TTS) system for Modern Standard Arabic which imposes a pitch contour on the output to indicate the force of the utterance (statement/query/command) and to mark emphasis (as specified by the use of non-canonical word orders). This TTS uses the diphone-based speech syn-thesiser Mbrola, for which you have to provide infor-mation about phone lengths. In the original implemen-tation of the system we hand-coded this information. The current paper describes an approach to estimat-ing phone lengths automatically by matching synthe-sised speech with human speech and adjusting the spec-ification of individual phone lengths accordingly.
Bookmarks Related papers MentionsView impact
Bookmarks Related papers MentionsView impact
aclweb.org
Bookmarks Related papers MentionsView impact
Computer Speech & Language, 2014
ABSTRACT This paper outlines a comprehensive system for au-tomatically generating a phonetic tran... more ABSTRACT This paper outlines a comprehensive system for au-tomatically generating a phonetic transcription of a given Arabic text which closely matches the pro-nunciation of the speakers. The presented system is based on a set of (language-dependent) pronunci-ation rules that works on converting fully diacriti-cised Arabic text into the actual sounds, along with a lexicon for exceptional words. This is a two-phase process: one-to-one grapheme to phoneme conver-sion and then phoneme-to-allophone conversion us-ing a set of “phonological rules”. Phonological rules operate on the phonemes and convert them to the actual sounds considering the neighboring phones or the containing syllable or word. This system is devel-oped for the purpose of delivering a robust Automatic Arabic Speech Recognition (AASR) system which is able to handle speech variation resulting from the mismatch between the text and the pronunciation. We anticipate that it could also be used for produc-ing natural sounding speech from an Arabic text-to-speech (ATTS) system as well, but we have not ex-tensively tested it in this application.
Bookmarks Related papers MentionsView impact
Computer Speech & Language, 2008
... b University of Alexandria, Egypt. ... Which of these gets chosen depends on the specific ten... more ... b University of Alexandria, Egypt. ... Which of these gets chosen depends on the specific tense markers that get added and on whether the verb is used in an active or apassive sentence. ... Other affixes cannot be settled so early. ...
Bookmarks Related papers MentionsView impact
Lecture Notes in Computer Science, 2006
Abstract. Syntactic analysis of Arabic poses two major problems. (i) Al-though the canonical orde... more Abstract. Syntactic analysis of Arabic poses two major problems. (i) Al-though the canonical order of Arabic sentences is VSO, a range of other orders are possible. In order to carry out such an analysis, then, it is neces-sary to have a grammatical framework and an associated ...
Bookmarks Related papers MentionsView impact
International Journal of Speech Technology
Bookmarks Related papers MentionsView impact
Proceedings of The 12th International Workshop on Semantic Evaluation
Bookmarks Related papers MentionsView impact
Information, Communication & Society
Bookmarks Related papers MentionsView impact
Proceedings of the Third Arabic Natural Language Processing Workshop
Bookmarks Related papers MentionsView impact
Bookmarks Related papers MentionsView impact
Information
Assigning sentiment labels to documents is, at first sight, a standard multi-label classification... more Assigning sentiment labels to documents is, at first sight, a standard multi-label classification task. Many approaches have been used for this task, but the current state-of-the-art solutions use deep neural networks (DNNs). As such, it seems likely that standard machine learning algorithms, such as these, will provide an effective approach. We describe an alternative approach, involving the use of probabilities to construct a weighted lexicon of sentiment terms, then modifying the lexicon and calculating optimal thresholds for each class. We show that this approach outperforms the use of DNNs and other standard algorithms. We believe that DNNs are not a universal panacea and that paying attention to the nature of the data that you are trying to learn from can be more important than trying out ever more powerful general purpose machine learning algorithms.
Bookmarks Related papers MentionsView impact
Computer Speech & Language, 2008
... b University of Alexandria, Egypt. ... Which of these gets chosen depends on the specific ten... more ... b University of Alexandria, Egypt. ... Which of these gets chosen depends on the specific tense markers that get added and on whether the verb is used in an active or apassive sentence. ... Other affixes cannot be settled so early. ...
Bookmarks Related papers MentionsView impact
Résumé - Abstract The work,described here is part of an attempt to provide a text-to-speech (TTS)... more Résumé - Abstract The work,described here is part of an attempt to provide a text-to-speech (TTS) system for
Bookmarks Related papers MentionsView impact
We have described elsewhere a text-to-speech (TTS) system for Modern Standard Arabic which impose... more We have described elsewhere a text-to-speech (TTS) system for Modern Standard Arabic which imposes a pitch contour on the output to indicate the force of the utterance (statement/query/command) and to mark emphasis (as specified by the use of non-canonical word orders). This TTS uses the diphone-based speech syn-thesiser Mbrola, for which you have to provide infor-mation about phone lengths. In the original implemen-tation of the system we hand-coded this information. The current paper describes an approach to estimat-ing phone lengths automatically by matching synthe-sised speech with human speech and adjusting the spec-ification of individual phone lengths accordingly.
Bookmarks Related papers MentionsView impact
Bookmarks Related papers MentionsView impact
aclweb.org
Bookmarks Related papers MentionsView impact
Computer Speech & Language, 2014
ABSTRACT This paper outlines a comprehensive system for au-tomatically generating a phonetic tran... more ABSTRACT This paper outlines a comprehensive system for au-tomatically generating a phonetic transcription of a given Arabic text which closely matches the pro-nunciation of the speakers. The presented system is based on a set of (language-dependent) pronunci-ation rules that works on converting fully diacriti-cised Arabic text into the actual sounds, along with a lexicon for exceptional words. This is a two-phase process: one-to-one grapheme to phoneme conver-sion and then phoneme-to-allophone conversion us-ing a set of “phonological rules”. Phonological rules operate on the phonemes and convert them to the actual sounds considering the neighboring phones or the containing syllable or word. This system is devel-oped for the purpose of delivering a robust Automatic Arabic Speech Recognition (AASR) system which is able to handle speech variation resulting from the mismatch between the text and the pronunciation. We anticipate that it could also be used for produc-ing natural sounding speech from an Arabic text-to-speech (ATTS) system as well, but we have not ex-tensively tested it in this application.
Bookmarks Related papers MentionsView impact
Computer Speech & Language, 2008
... b University of Alexandria, Egypt. ... Which of these gets chosen depends on the specific ten... more ... b University of Alexandria, Egypt. ... Which of these gets chosen depends on the specific tense markers that get added and on whether the verb is used in an active or apassive sentence. ... Other affixes cannot be settled so early. ...
Bookmarks Related papers MentionsView impact
Lecture Notes in Computer Science, 2006
Abstract. Syntactic analysis of Arabic poses two major problems. (i) Al-though the canonical orde... more Abstract. Syntactic analysis of Arabic poses two major problems. (i) Al-though the canonical order of Arabic sentences is VSO, a range of other orders are possible. In order to carry out such an analysis, then, it is neces-sary to have a grammatical framework and an associated ...
Bookmarks Related papers MentionsView impact