{"id":"https://openalex.org/W4308223422","doi":"https://doi.org/10.1145/3536220.3563689","title":"Towards Automatic Prediction of Non-Expert Perceived Speech Fluency Ratings","display_name":"Towards Automatic Prediction of Non-Expert Perceived Speech Fluency Ratings","publication_year":2022,"publication_date":"2022-11-04","ids":{"openalex":"https://openalex.org/W4308223422","doi":"https://doi.org/10.1145/3536220.3563689"},"language":"en","primary_location":{"id":"doi:10.1145/3536220.3563689","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3536220.3563689","pdf_url":null,"source":{"id":"https://openalex.org/S4363608440","display_name":"INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":false,"oa_status":"closed","oa_url":null,"any_repository_has_fulltext":false},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5008235303","display_name":"S. Pavankumar Dubagunta","orcid":"https://orcid.org/0000-0002-4307-8604"},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"S. Pavankumar Dubagunta","raw_affiliation_strings":["Uniphore Software Systems, India"],"raw_orcid":"https://orcid.org/0000-0002-4307-8604","affiliations":[{"raw_affiliation_string":"Uniphore Software Systems, India","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5111204141","display_name":"Edoardo Moneta","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Edoardo Moneta","raw_affiliation_strings":["speak and lunch S.A., Switzerland"],"raw_orcid":"https://orcid.org/0000-0002-8714-1409","affiliations":[{"raw_affiliation_string":"speak and lunch S.A., Switzerland","institution_ids":[]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5071802195","display_name":"Eleni Theocharopoulos","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Eleni Theocharopoulos","raw_affiliation_strings":["speak and lunch S.A., Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"speak and lunch S.A., Switzerland","institution_ids":[]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5043551083","display_name":"Mathew Magimai.-Doss","orcid":"https://orcid.org/0000-0002-8714-1409"},"institutions":[{"id":"https://openalex.org/I7495430","display_name":"Idiap Research Institute","ror":"https://ror.org/05932h694","country_code":"CH","type":"facility","lineage":["https://openalex.org/I7495430"]}],"countries":["CH"],"is_corresponding":false,"raw_author_name":"Mathew Magimai Doss","raw_affiliation_strings":["Idiap Research Institute, Switzerland"],"raw_orcid":null,"affiliations":[{"raw_affiliation_string":"Idiap Research Institute, Switzerland","institution_ids":["https://openalex.org/I7495430"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":1,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.3689,"has_fulltext":false,"cited_by_count":2,"citation_normalized_percentile":{"value":0.51981833,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":{"min":90,"max":95},"biblio":{"volume":null,"issue":null,"first_page":"7","last_page":"11"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.9995999932289124,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},"topics":[{"id":"https://openalex.org/T10860","display_name":"Speech and Audio Processing","score":0.9995999932289124,"subfield":{"id":"https://openalex.org/subfields/1711","display_name":"Signal Processing"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10201","display_name":"Speech Recognition and Synthesis","score":0.9975000023841858,"subfield":{"id":"https://openalex.org/subfields/1702","display_name":"Artificial Intelligence"},"field":{"id":"https://openalex.org/fields/17","display_name":"Computer Science"},"domain":{"id":"https://openalex.org/domains/3","display_name":"Physical Sciences"}},{"id":"https://openalex.org/T10667","display_name":"Emotion and Mood Recognition","score":0.9962000250816345,"subfield":{"id":"https://openalex.org/subfields/3205","display_name":"Experimental and Cognitive Psychology"},"field":{"id":"https://openalex.org/fields/32","display_name":"Psychology"},"domain":{"id":"https://openalex.org/domains/2","display_name":"Social Sciences"}}],"keywords":[{"id":"https://openalex.org/keywords/fluency","display_name":"Fluency","score":0.8341018557548523},{"id":"https://openalex.org/keywords/computer-science","display_name":"Computer science","score":0.6444100141525269},{"id":"https://openalex.org/keywords/speech-recognition","display_name":"Speech recognition","score":0.5868891477584839},{"id":"https://openalex.org/keywords/natural-language-processing","display_name":"Natural language processing","score":0.43746957182884216},{"id":"https://openalex.org/keywords/artificial-intelligence","display_name":"Artificial intelligence","score":0.41990673542022705},{"id":"https://openalex.org/keywords/psychology","display_name":"Psychology","score":0.2510807514190674},{"id":"https://openalex.org/keywords/mathematics-education","display_name":"Mathematics education","score":0.06258952617645264}],"concepts":[{"id":"https://openalex.org/C2777413886","wikidata":"https://www.wikidata.org/wiki/Q3276013","display_name":"Fluency","level":2,"score":0.8341018557548523},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.6444100141525269},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.5868891477584839},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.43746957182884216},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.41990673542022705},{"id":"https://openalex.org/C15744967","wikidata":"https://www.wikidata.org/wiki/Q9418","display_name":"Psychology","level":0,"score":0.2510807514190674},{"id":"https://openalex.org/C145420912","wikidata":"https://www.wikidata.org/wiki/Q853077","display_name":"Mathematics education","level":1,"score":0.06258952617645264}],"mesh":[],"locations_count":1,"locations":[{"id":"doi:10.1145/3536220.3563689","is_oa":false,"landing_page_url":"https://doi.org/10.1145/3536220.3563689","pdf_url":null,"source":{"id":"https://openalex.org/S4363608440","display_name":"INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":null,"host_organization_name":null,"host_organization_lineage":[],"host_organization_lineage_names":[],"type":"conference"},"license":"public-domain","license_id":"https://openalex.org/licenses/public-domain","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"INTERNATIONAL CONFERENCE ON MULTIMODAL INTERACTION","raw_type":"proceedings-article"}],"best_oa_location":null,"sustainable_development_goals":[{"id":"https://metadata.un.org/sdg/4","score":0.5400000214576721,"display_name":"Quality Education"}],"awards":[{"id":"https://openalex.org/G3159358282","display_name":null,"funder_award_id":"38843.1 IP-ICT","funder_id":"https://openalex.org/F4320327931","funder_display_name":"Innosuisse - Schweizerische Agentur f\u00fcr Innovationsf\u00f6rderung"},{"id":"https://openalex.org/G7595264094","display_name":null,"funder_award_id":"FLOSS 16036","funder_id":"https://openalex.org/F4320321942","funder_display_name":"Hasler Stiftung"}],"funders":[{"id":"https://openalex.org/F4320321942","display_name":"Hasler Stiftung","ror":"https://ror.org/04m3t9183"},{"id":"https://openalex.org/F4320327931","display_name":"Innosuisse - Schweizerische Agentur f\u00fcr Innovationsf\u00f6rderung","ror":"https://ror.org/05a2bhn71"}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":20,"referenced_works":["https://openalex.org/W1967625181","https://openalex.org/W2112844139","https://openalex.org/W2239141610","https://openalex.org/W2336514600","https://openalex.org/W2399733683","https://openalex.org/W2406311945","https://openalex.org/W2513345070","https://openalex.org/W2592641653","https://openalex.org/W2713293635","https://openalex.org/W2750506564","https://openalex.org/W2888859044","https://openalex.org/W2915722758","https://openalex.org/W2935807810","https://openalex.org/W2962736520","https://openalex.org/W2963175699","https://openalex.org/W2972763760","https://openalex.org/W2991435809","https://openalex.org/W2997591727","https://openalex.org/W3095248840","https://openalex.org/W4293571337"],"related_works":["https://openalex.org/W4391375266","https://openalex.org/W2748952813","https://openalex.org/W2365169615","https://openalex.org/W1970538215","https://openalex.org/W2400151637","https://openalex.org/W2975827637","https://openalex.org/W2354089692","https://openalex.org/W595497825","https://openalex.org/W4236323843","https://openalex.org/W3204019825"],"abstract_inverted_index":{"Automatic":[0],"speech":[1,93,103,141,155],"fluency":[2,31,54,94,156],"prediction":[3,89],"has":[4,78,97],"been":[5,99],"mainly":[6],"approached":[7],"from":[8,48,144],"the":[9,17,27,44,85,102,107,163,182],"perspective":[10],"of":[11,26,53,70,90,109,172,179],"computer":[12],"aided":[13],"language":[14],"learning,":[15],"where":[16,43],"system":[18,166],"tends":[19],"to":[20,24,66,106],"predict":[21],"ratings":[22,45,157],"similar":[23],"those":[25],"human":[28,184],"experts.":[29],"Speech":[30],"prediction,":[32],"however,":[33],"can":[34,158],"be":[35,159],"questioned":[36],"in":[37,101],"a":[38,76,80,168,175],"more":[39],"relaxed":[40],"social":[41,59],"setting,":[42],"arise":[46],"usually":[47],"non-experts;":[49],"indeed,":[50],"everyday":[51],"assessments":[52],"are":[55,68],"appraised":[56],"by":[57,149],"our":[58,110],"environment":[60],"and":[61,73,129,147,174],"encounters;":[62],"these":[63],"encounters":[64],"due":[65],"globalisation":[67],"becoming":[69],"international":[71],"nature":[72],"therefore":[74],"being":[75],"non-expert":[77,91,154],"become":[79],"norm.":[81],"This":[82],"paper":[83],"explores":[84],"latter":[86],"direction,":[87],"i.e.,":[88],"perceived":[92],"ratings,":[95],"which":[96],"not":[98],"studied":[100],"technology":[104],"literature,":[105],"best":[108,164],"knowledge.":[111],"Toward":[112],"that,":[113],"we":[114],"investigate":[115],"several":[116],"approaches,":[117],"namely,":[118],"(a)":[119],"low-level":[120],"descriptor":[121],"feature":[122],"functionals,":[123],"(b)":[124],"bag-of-audio":[125],"word":[126],"based":[127,133],"approach":[128],"(c)":[130],"neural":[131],"network":[132],"end-to-end":[134],"acoustic":[135],"modelling":[136],"approach.":[137],"Our":[138],"investigations":[139],"on":[140],"data":[142],"collected":[143],"54":[145],"speakers":[146],"rated":[148],"seven":[150],"non-experts":[151],"demonstrate":[152],"that":[153],"systematically":[160],"predicted,":[161],"with":[162,181],"performing":[165],"yielding":[167],"Pearson\u2019s":[169],"correlation":[170,177],"coefficient":[171,178],"0.66":[173],"Spearman\u2019s":[176],"0.67":[180],"median":[183],"scores.":[185]},"counts_by_year":[{"year":2025,"cited_by_count":1},{"year":2024,"cited_by_count":1}],"updated_date":"2026-07-30T17:31:21.811387","created_date":"2025-10-10T00:00:00"}