The (non)utility of linguistic features for predicting prominence in spontaneous speech
Conversational speech is characterized by prosodic variability which makes pitch accent prediction for this genre especially difficult. The linguistic literature points out that complex features such as information status, contrast and animacy help predict pitch accent placement. In this paper, we u...
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Tagungsbericht |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 57 |
|---|---|
| container_issue | |
| container_start_page | 54 |
| container_title | |
| container_volume | |
| creator | Brenier, J.M. Nenkova, A. Kothari, A. Whitton, L. Beaver, D. Jurafsky, D. |
| description | Conversational speech is characterized by prosodic variability which makes pitch accent prediction for this genre especially difficult. The linguistic literature points out that complex features such as information status, contrast and animacy help predict pitch accent placement. In this paper, we use a corpus annotated for such features to determine if they improve prominence prediction over traditional shallow features such as frequency and part-of-speech, or over new ones that we introduce. We demonstrate that while correlated with prominence, complex linguistic features do not improve prediction accuracy. Furthermore, the performance of our classifier is quite close to the ceiling defined by variability in human accent placement. An oracle experiment demonstrates, though, that at least some accuracy improvement is still possible. |
| doi_str_mv | 10.1109/SLT.2006.326815 |
| format | Conference Proceeding |
| fullrecord | <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_4123360</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>4123360</ieee_id><sourcerecordid>4123360</sourcerecordid><originalsourceid>FETCH-LOGICAL-i216t-6e909f2ff015c6235fd2cf61605627b1decaef0cb1ca4f49e0f97f8e4458bbd23</originalsourceid><addsrcrecordid>eNo1j01LxDAYhCMiqGvPHrzkqIfWfDVtjrL4BQUPVjwubfq-bqSblCY97L-3oM5lnmFgYAi55qzgnJn796YtBGO6kELXvDwhmalqroRSrK6kPCWX_0GU5ySL8ZutkkZrJi_IZ7sHeuuDv1uSG1060oB0dP5rcTE5SxG6tMwQKYaZTjMMzqa1XTEcnAdvgTpP4xR86jyEJa4MYPdX5Ay7MUL25xvy8fTYbl_y5u35dfvQ5E5wnXINhhkUiIyXVgtZ4iAsaq5ZqUXV8wFsB8hsz22nUBlgaCqsQamy7vtByA25-d11ALCbZnfo5uNOcSHl-u8HCgZTzw</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>The (non)utility of linguistic features for predicting prominence in spontaneous speech</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Brenier, J.M. ; Nenkova, A. ; Kothari, A. ; Whitton, L. ; Beaver, D. ; Jurafsky, D.</creator><creatorcontrib>Brenier, J.M. ; Nenkova, A. ; Kothari, A. ; Whitton, L. ; Beaver, D. ; Jurafsky, D.</creatorcontrib><description>Conversational speech is characterized by prosodic variability which makes pitch accent prediction for this genre especially difficult. The linguistic literature points out that complex features such as information status, contrast and animacy help predict pitch accent placement. In this paper, we use a corpus annotated for such features to determine if they improve prominence prediction over traditional shallow features such as frequency and part-of-speech, or over new ones that we introduce. We demonstrate that while correlated with prominence, complex linguistic features do not improve prediction accuracy. Furthermore, the performance of our classifier is quite close to the ceiling defined by variability in human accent placement. An oracle experiment demonstrates, though, that at least some accuracy improvement is still possible.</description><identifier>ISBN: 1424408725</identifier><identifier>ISBN: 9781424408726</identifier><identifier>EISBN: 9781424408733</identifier><identifier>EISBN: 1424408733</identifier><identifier>DOI: 10.1109/SLT.2006.326815</identifier><language>eng</language><subject>Accuracy ; Concrete ; Frequency ; Guidelines ; Humans ; Probability ; Robustness ; Speech synthesis</subject><ispartof>2006 IEEE Spoken Language Technology Workshop, 2006, p.54-57</ispartof><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/4123360$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,778,782,787,788,2054,27908,54903</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/4123360$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Brenier, J.M.</creatorcontrib><creatorcontrib>Nenkova, A.</creatorcontrib><creatorcontrib>Kothari, A.</creatorcontrib><creatorcontrib>Whitton, L.</creatorcontrib><creatorcontrib>Beaver, D.</creatorcontrib><creatorcontrib>Jurafsky, D.</creatorcontrib><title>The (non)utility of linguistic features for predicting prominence in spontaneous speech</title><title>2006 IEEE Spoken Language Technology Workshop</title><addtitle>SLT</addtitle><description>Conversational speech is characterized by prosodic variability which makes pitch accent prediction for this genre especially difficult. The linguistic literature points out that complex features such as information status, contrast and animacy help predict pitch accent placement. In this paper, we use a corpus annotated for such features to determine if they improve prominence prediction over traditional shallow features such as frequency and part-of-speech, or over new ones that we introduce. We demonstrate that while correlated with prominence, complex linguistic features do not improve prediction accuracy. Furthermore, the performance of our classifier is quite close to the ceiling defined by variability in human accent placement. An oracle experiment demonstrates, though, that at least some accuracy improvement is still possible.</description><subject>Accuracy</subject><subject>Concrete</subject><subject>Frequency</subject><subject>Guidelines</subject><subject>Humans</subject><subject>Probability</subject><subject>Robustness</subject><subject>Speech synthesis</subject><isbn>1424408725</isbn><isbn>9781424408726</isbn><isbn>9781424408733</isbn><isbn>1424408733</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2006</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><sourceid>RIE</sourceid><recordid>eNo1j01LxDAYhCMiqGvPHrzkqIfWfDVtjrL4BQUPVjwubfq-bqSblCY97L-3oM5lnmFgYAi55qzgnJn796YtBGO6kELXvDwhmalqroRSrK6kPCWX_0GU5ySL8ZutkkZrJi_IZ7sHeuuDv1uSG1060oB0dP5rcTE5SxG6tMwQKYaZTjMMzqa1XTEcnAdvgTpP4xR86jyEJa4MYPdX5Ay7MUL25xvy8fTYbl_y5u35dfvQ5E5wnXINhhkUiIyXVgtZ4iAsaq5ZqUXV8wFsB8hsz22nUBlgaCqsQamy7vtByA25-d11ALCbZnfo5uNOcSHl-u8HCgZTzw</recordid><startdate>20060101</startdate><enddate>20060101</enddate><creator>Brenier, J.M.</creator><creator>Nenkova, A.</creator><creator>Kothari, A.</creator><creator>Whitton, L.</creator><creator>Beaver, D.</creator><creator>Jurafsky, D.</creator><scope>6IE</scope><scope>6IL</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIL</scope></search><sort><creationdate>20060101</creationdate><title>The (non)utility of linguistic features for predicting prominence in spontaneous speech</title><author>Brenier, J.M. ; Nenkova, A. ; Kothari, A. ; Whitton, L. ; Beaver, D. ; Jurafsky, D.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i216t-6e909f2ff015c6235fd2cf61605627b1decaef0cb1ca4f49e0f97f8e4458bbd23</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2006</creationdate><topic>Accuracy</topic><topic>Concrete</topic><topic>Frequency</topic><topic>Guidelines</topic><topic>Humans</topic><topic>Probability</topic><topic>Robustness</topic><topic>Speech synthesis</topic><toplevel>online_resources</toplevel><creatorcontrib>Brenier, J.M.</creatorcontrib><creatorcontrib>Nenkova, A.</creatorcontrib><creatorcontrib>Kothari, A.</creatorcontrib><creatorcontrib>Whitton, L.</creatorcontrib><creatorcontrib>Beaver, D.</creatorcontrib><creatorcontrib>Jurafsky, D.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Electronic Library (IEL)</collection><collection>IEEE Proceedings Order Plans (POP All) 1998-Present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Brenier, J.M.</au><au>Nenkova, A.</au><au>Kothari, A.</au><au>Whitton, L.</au><au>Beaver, D.</au><au>Jurafsky, D.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>The (non)utility of linguistic features for predicting prominence in spontaneous speech</atitle><btitle>2006 IEEE Spoken Language Technology Workshop</btitle><stitle>SLT</stitle><date>2006-01-01</date><risdate>2006</risdate><spage>54</spage><epage>57</epage><pages>54-57</pages><isbn>1424408725</isbn><isbn>9781424408726</isbn><eisbn>9781424408733</eisbn><eisbn>1424408733</eisbn><abstract>Conversational speech is characterized by prosodic variability which makes pitch accent prediction for this genre especially difficult. The linguistic literature points out that complex features such as information status, contrast and animacy help predict pitch accent placement. In this paper, we use a corpus annotated for such features to determine if they improve prominence prediction over traditional shallow features such as frequency and part-of-speech, or over new ones that we introduce. We demonstrate that while correlated with prominence, complex linguistic features do not improve prediction accuracy. Furthermore, the performance of our classifier is quite close to the ceiling defined by variability in human accent placement. An oracle experiment demonstrates, though, that at least some accuracy improvement is still possible.</abstract><doi>10.1109/SLT.2006.326815</doi><tpages>4</tpages><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISBN: 1424408725 |
| ispartof | 2006 IEEE Spoken Language Technology Workshop, 2006, p.54-57 |
| issn | |
| language | eng |
| recordid | cdi_ieee_primary_4123360 |
| source | IEEE Electronic Library (IEL) Conference Proceedings |
| subjects | Accuracy Concrete Frequency Guidelines Humans Probability Robustness Speech synthesis |
| title | The (non)utility of linguistic features for predicting prominence in spontaneous speech |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-16T20%3A16%3A45IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_6IE&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=The%20(non)utility%20of%20linguistic%20features%20for%20predicting%20prominence%20in%20spontaneous%20speech&rft.btitle=2006%20IEEE%20Spoken%20Language%20Technology%20Workshop&rft.au=Brenier,%20J.M.&rft.date=2006-01-01&rft.spage=54&rft.epage=57&rft.pages=54-57&rft.isbn=1424408725&rft.isbn_list=9781424408726&rft_id=info:doi/10.1109/SLT.2006.326815&rft_dat=%3Cieee_6IE%3E4123360%3C/ieee_6IE%3E%3Curl%3E%3C/url%3E&rft.eisbn=9781424408733&rft.eisbn_list=1424408733&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=4123360&rfr_iscdi=true |