{"id":"https://openalex.org/W7147218531","doi":"https://doi.org/10.1109/access.2026.3679735","title":"Efficient Transformer Models via Language-Aware Frequency-Based Vocabulary Pruning","display_name":"Efficient Transformer Models via Language-Aware Frequency-Based Vocabulary Pruning","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7147218531","doi":"https://doi.org/10.1109/access.2026.3679735"},"language":"en","primary_location":{"id":"doi:10.1109/access.2026.3679735","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2026.3679735","pdf_url":null,"source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Access","raw_type":"journal-article"},"type":"article","indexed_in":["crossref","doaj"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.1109/access.2026.3679735","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5010880387","display_name":"Mukhlish Fuadi","orcid":"https://orcid.org/0000-0002-7595-7646"},"institutions":[{"id":"https://openalex.org/I166843116","display_name":"Sepuluh Nopember Institute of Technology","ror":"https://ror.org/05kbmmt89","country_code":"ID","type":"education","lineage":["https://openalex.org/I166843116"]}],"countries":["ID"],"is_corresponding":false,"raw_author_name":"Mukhlish Fuadi","raw_affiliation_strings":["1Department of Electrical Engineering, Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia"],"raw_orcid":"https://orcid.org/0000-0002-7595-7646","affiliations":[{"raw_affiliation_string":"1Department of Electrical Engineering, Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia","institution_ids":["https://openalex.org/I166843116"]}]},{"author_position":"middle","author":{"id":"https://openalex.org/A5078051957","display_name":"Adhi Dharma Wibawa","orcid":"https://orcid.org/0000-0002-0651-5078"},"institutions":[{"id":"https://openalex.org/I166843116","display_name":"Sepuluh Nopember Institute of Technology","ror":"https://ror.org/05kbmmt89","country_code":"ID","type":"education","lineage":["https://openalex.org/I166843116"]}],"countries":["ID"],"is_corresponding":false,"raw_author_name":"Adhi Dharma Wibawa","raw_affiliation_strings":["1Department of Electrical Engineering, Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia"],"raw_orcid":"https://orcid.org/0000-0002-0651-5078","affiliations":[{"raw_affiliation_string":"1Department of Electrical Engineering, Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia","institution_ids":["https://openalex.org/I166843116"]}]},{"author_position":"last","author":{"id":"https://openalex.org/A5019052119","display_name":"Surya Sumpeno","orcid":"https://orcid.org/0000-0002-1744-1342"},"institutions":[{"id":"https://openalex.org/I166843116","display_name":"Sepuluh Nopember Institute of Technology","ror":"https://ror.org/05kbmmt89","country_code":"ID","type":"education","lineage":["https://openalex.org/I166843116"]}],"countries":["ID"],"is_corresponding":false,"raw_author_name":"Surya Sumpeno","raw_affiliation_strings":["1Department of Electrical Engineering, Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia"],"raw_orcid":"https://orcid.org/0000-0002-1744-1342","affiliations":[{"raw_affiliation_string":"1Department of Electrical Engineering, Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia","institution_ids":["https://openalex.org/I166843116"]}]}],"institutions":[],"countries_distinct_count":1,"institutions_distinct_count":3,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":{"value":1850,"currency":"USD","value_usd":1850},"apc_paid":{"value":1850,"currency":"USD","value_usd":1850},"fwci":19.5563,"has_fulltext":false,"cited_by_count":1,"citation_normalized_percentile":{"value":0.98893423,"is_in_top_1_percent":false,"is_in_top_10_percent":true},"cited_by_percentile_year":{"min":95,"max":98},"biblio":{"volume":"14","issue":null,"first_page":"50993","last_page":"51006"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.6208000183105469,"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"}},"topics":[{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.6208000183105469,"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/T10664","display_name":"Sentiment Analysis and Opinion Mining","score":0.10109999775886536,"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/T10181","display_name":"Natural Language Processing Techniques","score":0.047200001776218414,"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"}}],"keywords":[{"id":"https://openalex.org/keywords/inference","display_name":"Inference","score":0.5812000036239624},{"id":"https://openalex.org/keywords/transformer","display_name":"Transformer","score":0.5331000089645386},{"id":"https://openalex.org/keywords/vocabulary","display_name":"Vocabulary","score":0.4957999885082245},{"id":"https://openalex.org/keywords/lexical-analysis","display_name":"Lexical analysis","score":0.4153999984264374},{"id":"https://openalex.org/keywords/word","display_name":"Word (group theory)","score":0.314300000667572},{"id":"https://openalex.org/keywords/keyword-spotting","display_name":"Keyword spotting","score":0.3131999969482422},{"id":"https://openalex.org/keywords/word-embedding","display_name":"Word embedding","score":0.3127000033855438},{"id":"https://openalex.org/keywords/benchmarking","display_name":"Benchmarking","score":0.3093000054359436}],"concepts":[{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.8348000049591064},{"id":"https://openalex.org/C2776214188","wikidata":"https://www.wikidata.org/wiki/Q408386","display_name":"Inference","level":2,"score":0.5812000036239624},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.5514000058174133},{"id":"https://openalex.org/C66322947","wikidata":"https://www.wikidata.org/wiki/Q11658","display_name":"Transformer","level":3,"score":0.5331000089645386},{"id":"https://openalex.org/C2777601683","wikidata":"https://www.wikidata.org/wiki/Q6499736","display_name":"Vocabulary","level":2,"score":0.4957999885082245},{"id":"https://openalex.org/C28490314","wikidata":"https://www.wikidata.org/wiki/Q189436","display_name":"Speech recognition","level":1,"score":0.4323999881744385},{"id":"https://openalex.org/C176982825","wikidata":"https://www.wikidata.org/wiki/Q835922","display_name":"Lexical analysis","level":2,"score":0.4153999984264374},{"id":"https://openalex.org/C119857082","wikidata":"https://www.wikidata.org/wiki/Q2539","display_name":"Machine learning","level":1,"score":0.3628000020980835},{"id":"https://openalex.org/C204321447","wikidata":"https://www.wikidata.org/wiki/Q30642","display_name":"Natural language processing","level":1,"score":0.3357999920845032},{"id":"https://openalex.org/C90805587","wikidata":"https://www.wikidata.org/wiki/Q10944557","display_name":"Word (group theory)","level":2,"score":0.314300000667572},{"id":"https://openalex.org/C2781213101","wikidata":"https://www.wikidata.org/wiki/Q6398558","display_name":"Keyword spotting","level":2,"score":0.3131999969482422},{"id":"https://openalex.org/C2777462759","wikidata":"https://www.wikidata.org/wiki/Q18395344","display_name":"Word embedding","level":3,"score":0.3127000033855438},{"id":"https://openalex.org/C86251818","wikidata":"https://www.wikidata.org/wiki/Q816754","display_name":"Benchmarking","level":2,"score":0.3093000054359436},{"id":"https://openalex.org/C39927690","wikidata":"https://www.wikidata.org/wiki/Q11197","display_name":"Logarithm","level":2,"score":0.30799999833106995},{"id":"https://openalex.org/C2778869765","wikidata":"https://www.wikidata.org/wiki/Q6028363","display_name":"Inefficiency","level":2,"score":0.3061000108718872},{"id":"https://openalex.org/C137293760","wikidata":"https://www.wikidata.org/wiki/Q3621696","display_name":"Language model","level":2,"score":0.28999999165534973},{"id":"https://openalex.org/C82876162","wikidata":"https://www.wikidata.org/wiki/Q17096504","display_name":"Latency (audio)","level":2,"score":0.26759999990463257},{"id":"https://openalex.org/C97541855","wikidata":"https://www.wikidata.org/wiki/Q830687","display_name":"Reinforcement learning","level":2,"score":0.263700008392334},{"id":"https://openalex.org/C108010975","wikidata":"https://www.wikidata.org/wiki/Q500094","display_name":"Pruning","level":2,"score":0.26339998841285706},{"id":"https://openalex.org/C49740808","wikidata":"https://www.wikidata.org/wiki/Q7550116","display_name":"Sobol sequence","level":3,"score":0.25609999895095825},{"id":"https://openalex.org/C114289077","wikidata":"https://www.wikidata.org/wiki/Q3284399","display_name":"Statistical model","level":2,"score":0.2508000135421753}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.1109/access.2026.3679735","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2026.3679735","pdf_url":null,"source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Access","raw_type":"journal-article"},{"id":"pmh:oai:doaj.org/article:f5ed12b3d4ae4b4282c2290ffd1e606d","is_oa":true,"landing_page_url":"https://doaj.org/article/f5ed12b3d4ae4b4282c2290ffd1e606d","pdf_url":null,"source":{"id":"https://openalex.org/S4306401280","display_name":"DOAJ (DOAJ: Directory of Open Access Journals)","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":"repository"},"license":"cc-by-sa","license_id":"https://openalex.org/licenses/cc-by-sa","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":"IEEE Access, Vol 14, Pp 50993-51006 (2026)","raw_type":"article"}],"best_oa_location":{"id":"doi:10.1109/access.2026.3679735","is_oa":true,"landing_page_url":"https://doi.org/10.1109/access.2026.3679735","pdf_url":null,"source":{"id":"https://openalex.org/S2485537415","display_name":"IEEE Access","issn_l":"2169-3536","issn":["2169-3536"],"is_oa":true,"is_in_doaj":true,"is_core":true,"host_organization":"https://openalex.org/P4310319808","host_organization_name":"Institute of Electrical and Electronics Engineers","host_organization_lineage":["https://openalex.org/P4310319808"],"host_organization_lineage_names":["Institute of Electrical and Electronics Engineers"],"type":"journal"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"IEEE Access","raw_type":"journal-article"},"sustainable_development_goals":[],"awards":[{"id":"https://openalex.org/G5356923722","display_name":null,"funder_award_id":"LOG-9715/LPDP/LPDP.3/2024","funder_id":"https://openalex.org/F4320328515","funder_display_name":"Lembaga Pengelola Dana Pendidikan"}],"funders":[{"id":"https://openalex.org/F4320328515","display_name":"Lembaga Pengelola Dana Pendidikan","ror":null}],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Multilingual":[0],"Transformer":[1,208],"models":[2,209],"offer":[3],"effective":[4],"cross-lingual":[5,135],"generalization":[6],"capabilities.":[7],"However,":[8],"their":[9],"architecture":[10],"suffers":[11],"from":[12,94],"embedding-parameter":[13],"overhead":[14],"due":[15],"to":[16,43,84,97,133,176],"massive":[17],"vocabulary":[18,35,65,83],"sizes,":[19],"where":[20],"most":[21],"embedding":[22],"parameters":[23],"are":[24],"redundant":[25],"for":[26,206],"specific":[27],"monolingual":[28,60],"tasks.":[29,76],"This":[30],"study":[31],"investigates":[32],"language-aware":[33],"frequency-based":[34],"pruning":[36],"as":[37,188,202],"a":[38,147,189,203],"deterministic":[39],"and":[40,62,72,106,121,141],"reproducible":[41],"strategy":[42],"mitigate":[44],"this":[45,199],"inefficiency":[46],"without":[47],"architectural":[48],"modifications":[49],"or":[50],"additional":[51],"large-scale":[52],"pre-training.":[53],"We":[54],"evaluate":[55],"the":[56,82,129,162,177,193,196],"mDeBERTa-v3-base":[57],"model":[58],"under":[59],"(Indonesian)":[61],"hybrid":[63,130,163],"(Indonesian\u2013English)":[64],"configurations":[66],"on":[67,166,186],"Named":[68],"Entity":[69],"Recognition":[70],"(NER)":[71,120],"Sentiment":[73],"Analysis":[74],"(SA)":[75,124],"Multi-seed":[77],"experiments":[78],"show":[79],"that":[80,161],"reducing":[81],"30k":[85],"tokens":[86],"lowers":[87],"peak":[88],"GPU":[89],"memory":[90],"usage":[91],"during":[92],"inference":[93,104],"6.64":[95],"GB":[96,99],"4.70":[98],"(\u224829%),":[100],"while":[101],"maintaining":[102],"comparable":[103],"latency":[105],"minimal":[107,154],"performance":[108,165],"degradation,":[109],"with":[110],"an":[111],"average":[112],"drop":[113],"of":[114,150,198],"less":[115],"than":[116],"1.5%":[117],"across":[118],"entity-level":[119],"sentence-level":[122,167],"classification":[123,168],"F1-scores.":[125],"Further":[126],"analysis":[127],"attributes":[128],"model\u2019s":[131,164],"stability":[132],"substantial":[134],"subword":[136,148],"overlap":[137],"(over":[138],"24,000":[139],"tokens)":[140],"efficient":[142],"tokenization":[143],"behavior,":[144],"reflected":[145],"in":[146,210],"fertility":[149],"approximately":[151],"1.008,":[152],"indicating":[153],"word":[155],"fragmentation.":[156],"Statistical":[157],"significance":[158],"testing":[159],"confirms":[160],"tasks":[169],"shows":[170],"no":[171],"statistically":[172],"significant":[173],"difference":[174],"compared":[175],"baseline":[178],"(<italic":[179],"xmlns:mml=\"http://www.w3.org/1998/Math/MathML\"":[180],"xmlns:xlink=\"http://www.w3.org/1999/xlink\">p</i>":[181],">":[182],"0.05).":[183],"Although":[184],"evaluated":[185],"Indonesian":[187],"representative":[190],"case":[191],"study,":[192],"results":[194],"illustrate":[195],"potential":[197],"language-agnostic":[200],"framework":[201],"practical":[204],"solution":[205],"optimizing":[207],"resource-constrained":[211],"deployment":[212],"scenarios.":[213]},"counts_by_year":[{"year":2026,"cited_by_count":1}],"updated_date":"2026-06-11T09:08:48.828518","created_date":"2026-04-02T00:00:00"}