{"paper":{"title":"Harf-Speech: A Clinically Aligned Framework for Arabic Phoneme-Level Speech Assessment","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Harf-Speech scores Arabic pronunciation at the phoneme level with 0.791 correlation to expert pathologists.","cross_cats":["cs.AI","cs.CL","cs.SD"],"primary_cat":"eess.AS","authors_text":"Abdulrhman Aljouie, Asif Azad, Ayah Othman Sindi, Bdour Alwuqaysi, Ehsan Hoque, MD Sadik Hossain Shanto, Mohammad Sadat Hossain, Sabri Boughorbel, Yahya Bokhari","submitted_at":"2026-03-11T04:10:35Z","abstract_excerpt":"Automated phoneme-level pronunciation assessment is vital for scalable speech therapy and language learning, yet validated tools for Arabic remain scarce. We present Harf-Speech, a modular system scoring Arabic pronunciation at the phoneme level on a clinical scale. It combines an MSA phonetizer, a fine-tuned speech-to-phoneme model, Levenshtein alignment, and a blended scorer using longest common subsequence and edit-distance metrics. We fine-tune three ASR architectures on Arabic phoneme data and benchmark them with zero-shot multimodal models; the best, OmniASR-CTC-1B-v2, achieves 8.92% pho"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Harf-Speech attains a Pearson correlation of 0.791 and ICC(2,1) of 0.659 with mean expert scores, outperforming existing end-to-end assessment frameworks.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The 40 utterances scored by three pathologists are representative of the clinical population and that the blended LCS-plus-edit-distance metric faithfully captures the clinical judgment used by experts.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Harf-Speech delivers phoneme-level Arabic pronunciation scores that correlate 0.79 with certified speech-language pathologists on 40 utterances.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Harf-Speech scores Arabic pronunciation at the phoneme level with 0.791 correlation to expert pathologists.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"49c329ded3649b5fee6b73e115b2646794f663a6b46d5e5bef41489b010ec591"},"source":{"id":"2604.06191","kind":"arxiv","version":2},"verdict":{"id":"57a87f6b-1df2-4115-a99a-a41f29e8692d","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T13:37:38.480667Z","strongest_claim":"Harf-Speech attains a Pearson correlation of 0.791 and ICC(2,1) of 0.659 with mean expert scores, outperforming existing end-to-end assessment frameworks.","one_line_summary":"Harf-Speech delivers phoneme-level Arabic pronunciation scores that correlate 0.79 with certified speech-language pathologists on 40 utterances.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The 40 utterances scored by three pathologists are representative of the clinical population and that the blended LCS-plus-edit-distance metric faithfully captures the clinical judgment used by experts.","pith_extraction_headline":"Harf-Speech scores Arabic pronunciation at the phoneme level with 0.791 correlation to expert pathologists."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.06191/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"6af27d2f2e76041dab127244f663de9259ffb1c0df052de5b323f80f26e4802d"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}