{"id":"https://openalex.org/W7166880637","doi":"https://doi.org/10.18653/v1/2026.acl-long.18","title":"MINTQA: A Multi-Hop Question Answering Benchmark for Evaluating LLMs on New and Long-tail Knowledge","display_name":"MINTQA: A Multi-Hop Question Answering Benchmark for Evaluating LLMs on New and Long-tail Knowledge","publication_year":2026,"publication_date":"2026-01-01","ids":{"openalex":"https://openalex.org/W7166880637","doi":"https://doi.org/10.18653/v1/2026.acl-long.18"},"language":"en","primary_location":{"id":"doi:10.18653/v1/2026.acl-long.18","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.18","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"type":"conference-paper","indexed_in":["crossref"],"open_access":{"is_oa":true,"oa_status":"gold","oa_url":"https://doi.org/10.18653/v1/2026.acl-long.18","any_repository_has_fulltext":true},"authorships":[{"author_position":"first","author":{"id":"https://openalex.org/A5139843943","display_name":"Jie He","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jie He","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139803362","display_name":"Nan Hu","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Nan Hu","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5000467097","display_name":"Wanqiu Long","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Wanqiu Long","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"middle","author":{"id":"https://openalex.org/A5139753514","display_name":"Jiaoyan Chen","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jiaoyan Chen","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]},{"author_position":"last","author":{"id":"https://openalex.org/A5139849912","display_name":"Jeff Z. Pan","orcid":null},"institutions":[],"countries":[],"is_corresponding":false,"raw_author_name":"Jeff Z. Pan","raw_affiliation_strings":[],"raw_orcid":null,"affiliations":[]}],"institutions":[],"countries_distinct_count":0,"institutions_distinct_count":0,"corresponding_author_ids":[],"corresponding_institution_ids":[],"apc_list":null,"apc_paid":null,"fwci":0.0,"has_fulltext":false,"cited_by_count":0,"citation_normalized_percentile":{"value":0.86889563,"is_in_top_1_percent":false,"is_in_top_10_percent":false},"cited_by_percentile_year":null,"biblio":{"volume":null,"issue":null,"first_page":"445","last_page":"479"},"is_retracted":false,"is_paratext":false,"is_xpac":false,"primary_topic":{"id":"https://openalex.org/T10028","display_name":"Topic Modeling","score":0.6347000002861023,"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":0.6347000002861023,"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":0.06530000269412994,"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/T13274","display_name":"Expert finding and Q&A systems","score":0.052000001072883606,"subfield":{"id":"https://openalex.org/subfields/1710","display_name":"Information Systems"},"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/question-answering","display_name":"Question answering","score":0.6996999979019165},{"id":"https://openalex.org/keywords/benchmark","display_name":"Benchmark (surveying)","score":0.4851999878883362},{"id":"https://openalex.org/keywords/knowledge-based-systems","display_name":"Knowledge-based systems","score":0.29989999532699585},{"id":"https://openalex.org/keywords/key","display_name":"Key (lock)","score":0.2962999939918518},{"id":"https://openalex.org/keywords/knowledge-base","display_name":"Knowledge base","score":0.28790000081062317},{"id":"https://openalex.org/keywords/questions-and-answers","display_name":"Questions and answers","score":0.28189998865127563}],"concepts":[{"id":"https://openalex.org/C44291984","wikidata":"https://www.wikidata.org/wiki/Q1074173","display_name":"Question answering","level":2,"score":0.6996999979019165},{"id":"https://openalex.org/C41008148","wikidata":"https://www.wikidata.org/wiki/Q21198","display_name":"Computer science","level":0,"score":0.5115000009536743},{"id":"https://openalex.org/C185798385","wikidata":"https://www.wikidata.org/wiki/Q1161707","display_name":"Benchmark (surveying)","level":2,"score":0.4851999878883362},{"id":"https://openalex.org/C56739046","wikidata":"https://www.wikidata.org/wiki/Q192060","display_name":"Knowledge management","level":1,"score":0.4034000039100647},{"id":"https://openalex.org/C154945302","wikidata":"https://www.wikidata.org/wiki/Q11660","display_name":"Artificial intelligence","level":1,"score":0.33709999918937683},{"id":"https://openalex.org/C2522767166","wikidata":"https://www.wikidata.org/wiki/Q2374463","display_name":"Data science","level":1,"score":0.33219999074935913},{"id":"https://openalex.org/C23123220","wikidata":"https://www.wikidata.org/wiki/Q816826","display_name":"Information retrieval","level":1,"score":0.30820000171661377},{"id":"https://openalex.org/C115925183","wikidata":"https://www.wikidata.org/wiki/Q1412694","display_name":"Knowledge-based systems","level":2,"score":0.29989999532699585},{"id":"https://openalex.org/C26517878","wikidata":"https://www.wikidata.org/wiki/Q228039","display_name":"Key (lock)","level":2,"score":0.2962999939918518},{"id":"https://openalex.org/C4554734","wikidata":"https://www.wikidata.org/wiki/Q593744","display_name":"Knowledge base","level":2,"score":0.28790000081062317},{"id":"https://openalex.org/C3019144022","wikidata":"https://www.wikidata.org/wiki/Q4124998","display_name":"Questions and answers","level":2,"score":0.28189998865127563},{"id":"https://openalex.org/C189430467","wikidata":"https://www.wikidata.org/wiki/Q7293293","display_name":"Ranking (information retrieval)","level":2,"score":0.2777999937534332},{"id":"https://openalex.org/C62360110","wikidata":"https://www.wikidata.org/wiki/Q96777007","display_name":"Circumscription","level":2,"score":0.26019999384880066},{"id":"https://openalex.org/C180198813","wikidata":"https://www.wikidata.org/wiki/Q121182","display_name":"Information system","level":2,"score":0.2554999887943268},{"id":"https://openalex.org/C2780791683","wikidata":"https://www.wikidata.org/wiki/Q846785","display_name":"Action (physics)","level":2,"score":0.250900000333786}],"mesh":[],"locations_count":2,"locations":[{"id":"doi:10.18653/v1/2026.acl-long.18","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.18","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},{"id":"pmh:oai:pure.ed.ac.uk:openaire/0dc21762-5181-4ffe-8f94-aa6fed608569","is_oa":true,"landing_page_url":"https://www.research.ed.ac.uk/en/publications/0dc21762-5181-4ffe-8f94-aa6fed608569","pdf_url":null,"source":{"id":"https://openalex.org/S4306400321","display_name":"Edinburgh Research Explorer (University of Edinburgh)","issn_l":null,"issn":null,"is_oa":false,"is_in_doaj":false,"is_core":false,"host_organization":"https://openalex.org/I98677209","host_organization_name":"University of Edinburgh","host_organization_lineage":["https://openalex.org/I98677209"],"host_organization_lineage_names":[],"type":"repository"},"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"submittedVersion","is_accepted":false,"is_published":false,"raw_source_name":null,"raw_type":"info:eu-repo/semantics/conferenceObject"}],"best_oa_location":{"id":"doi:10.18653/v1/2026.acl-long.18","is_oa":true,"landing_page_url":"https://doi.org/10.18653/v1/2026.acl-long.18","pdf_url":null,"source":null,"license":"cc-by","license_id":"https://openalex.org/licenses/cc-by","version":"publishedVersion","is_accepted":true,"is_published":true,"raw_source_name":"Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)","raw_type":"proceedings-article"},"sustainable_development_goals":[],"awards":[],"funders":[],"has_content":{"pdf":false,"grobid_xml":false},"content_urls":null,"referenced_works_count":0,"referenced_works":[],"related_works":[],"abstract_inverted_index":{"Retrieval-Augmented":[0],"Generation":[1],"(RAG)":[2],"enhances":[3],"large":[4],"language":[5],"models":[6],"(LLMs)":[7],"by":[8],"integrating":[9],"external":[10],"knowledge,":[11,82],"enabling":[12],"them":[13],"to":[14,61,105],"tackle":[15],"knowledge-intensive":[16],"tasks.":[17],"However,":[18],"limited":[19],"research":[20],"has":[21],"explored":[22],"how":[23,144],"LLMs":[24,101,121,145],"effectively":[25],"leverage":[26],"RAG":[27,163],"techniques":[28],"for":[29,72,79],"multi-hop":[30,63,97,135,157],"question":[31,85],"answering":[32],"(QA),":[33],"particularly":[34],"when":[35,160],"handling":[36],"knowledge":[37,75,107,129,147],"with":[38,39,83,87,108,148,162],"varying":[40,109],"degrees":[41],"of":[42,100,111],"familiarity.":[43],"In":[44],"this":[45],"paper,":[46],"we":[47],"introduce":[48],"MINTQA":[49],"(Multi-hop":[50],"Question":[51],"Answering":[52],"on":[53,66],"New":[54],"and":[55,76,90,102,134,151],"Tail":[56],"Knowledge),":[57],"a":[58],"benchmark":[59,93],"designed":[60],"evaluate":[62,118],"QA":[64,125],"performance":[65],"questions":[67],"involving":[68],"10,479":[69],"question-answer":[70],"pairs":[71,78],"evaluating":[73],"old/new":[74],"17,887":[77],"assessing":[80],"popular/unpopular":[81],"each":[84],"equipped":[86],"corresponding":[88],"sub-questions":[89],"answers.":[91],"This":[92],"primarily":[94],"evaluates":[95],"the":[96,114],"reasoning":[98,115,158],"ability":[99],"their":[103,156],"capacity":[104],"handle":[106,146],"levels":[110],"familiarity":[112,150],"during":[113],"process.":[116],"We":[117],"22":[119],"state-of-the-art":[120],"using":[122],"three":[123],"distinct":[124],"strategies:":[126],"LLM-based":[127],"parameterized":[128],"QA,":[130,133],"direct":[131],"RAG-enhanced":[132,136],"QA.":[137],"Our":[138],"experiments":[139],"reveal":[140],"key":[141],"challenges":[142],"in":[143],"different":[149],"offer":[152],"insights":[153],"into":[154],"improving":[155],"capabilities":[159],"combined":[161],"techniques.":[164]},"counts_by_year":[],"updated_date":"2026-07-29T14:22:42.915294","created_date":"2026-07-02T00:00:00"}