{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:HQQNQWHDORSYWL2EIP6YT2DDEP","short_pith_number":"pith:HQQNQWHD","schema_version":"1.0","canonical_sha256":"3c20d858e374658b2f4443fd89e86323da7aac2409c2f818b3b8db470be8dda7","source":{"kind":"arxiv","id":"2508.04206","version":1},"attestation_state":"computed","paper":{"title":"ViLLA-MMBench: A Unified Benchmark Suite for LLM-Augmented Multimodal Movie Recommendation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Ali Tourani, Fatemeh Nazary, Tommaso Di Noia, Yashar Deldjoo","submitted_at":"2025-08-06T08:39:07Z","abstract_excerpt":"Recommending long-form video content demands joint modeling of visual, audio, and textual modalities, yet most benchmarks address only raw features or narrow fusion. We present ViLLA-MMBench, a reproducible, extensible benchmark for LLM-augmented multimodal movie recommendation. Built on MovieLens and MMTF-14K, it aligns dense item embeddings from three modalities: audio (block-level, i-vector), visual (CNN, AVF), and text. Missing or sparse metadata is automatically enriched using state-of-the-art LLMs (e.g., OpenAI Ada), generating high-quality synopses for thousands of movies. All text (raw"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2508.04206","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.IR","submitted_at":"2025-08-06T08:39:07Z","cross_cats_sorted":[],"title_canon_sha256":"e1993349a11b56c30936faf3a57346f78935a4413035fe176e3c257696e209fb","abstract_canon_sha256":"ad031f05253cef25c0c466f408aaa119fe9427670114945c2f3891d45eff1e60"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:49:22.777408Z","signature_b64":"nxmRZ0u34g6JN7/gq0iRTZSB5Va59SoTQD1nwHfUKBomdYVfOoLGS9BwWKGgpATWHxvaKjFxpp08ro6WYmfBDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3c20d858e374658b2f4443fd89e86323da7aac2409c2f818b3b8db470be8dda7","last_reissued_at":"2026-07-05T11:49:22.776917Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:49:22.776917Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ViLLA-MMBench: A Unified Benchmark Suite for LLM-Augmented Multimodal Movie Recommendation","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.IR","authors_text":"Ali Tourani, Fatemeh Nazary, Tommaso Di Noia, Yashar Deldjoo","submitted_at":"2025-08-06T08:39:07Z","abstract_excerpt":"Recommending long-form video content demands joint modeling of visual, audio, and textual modalities, yet most benchmarks address only raw features or narrow fusion. We present ViLLA-MMBench, a reproducible, extensible benchmark for LLM-augmented multimodal movie recommendation. Built on MovieLens and MMTF-14K, it aligns dense item embeddings from three modalities: audio (block-level, i-vector), visual (CNN, AVF), and text. Missing or sparse metadata is automatically enriched using state-of-the-art LLMs (e.g., OpenAI Ada), generating high-quality synopses for thousands of movies. All text (raw"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.04206","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2508.04206/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":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2508.04206","created_at":"2026-07-05T11:49:22.776978+00:00"},{"alias_kind":"arxiv_version","alias_value":"2508.04206v1","created_at":"2026-07-05T11:49:22.776978+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.04206","created_at":"2026-07-05T11:49:22.776978+00:00"},{"alias_kind":"pith_short_12","alias_value":"HQQNQWHDORSY","created_at":"2026-07-05T11:49:22.776978+00:00"},{"alias_kind":"pith_short_16","alias_value":"HQQNQWHDORSYWL2E","created_at":"2026-07-05T11:49:22.776978+00:00"},{"alias_kind":"pith_short_8","alias_value":"HQQNQWHD","created_at":"2026-07-05T11:49:22.776978+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.09595","citing_title":"Popcorn: A Configurable Benchmark for Visual Evidence in Multimodal Movie Recommendation","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HQQNQWHDORSYWL2EIP6YT2DDEP","json":"https://pith.science/pith/HQQNQWHDORSYWL2EIP6YT2DDEP.json","graph_json":"https://pith.science/api/pith-number/HQQNQWHDORSYWL2EIP6YT2DDEP/graph.json","events_json":"https://pith.science/api/pith-number/HQQNQWHDORSYWL2EIP6YT2DDEP/events.json","paper":"https://pith.science/paper/HQQNQWHD"},"agent_actions":{"view_html":"https://pith.science/pith/HQQNQWHDORSYWL2EIP6YT2DDEP","download_json":"https://pith.science/pith/HQQNQWHDORSYWL2EIP6YT2DDEP.json","view_paper":"https://pith.science/paper/HQQNQWHD","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2508.04206&json=true","fetch_graph":"https://pith.science/api/pith-number/HQQNQWHDORSYWL2EIP6YT2DDEP/graph.json","fetch_events":"https://pith.science/api/pith-number/HQQNQWHDORSYWL2EIP6YT2DDEP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HQQNQWHDORSYWL2EIP6YT2DDEP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HQQNQWHDORSYWL2EIP6YT2DDEP/action/storage_attestation","attest_author":"https://pith.science/pith/HQQNQWHDORSYWL2EIP6YT2DDEP/action/author_attestation","sign_citation":"https://pith.science/pith/HQQNQWHDORSYWL2EIP6YT2DDEP/action/citation_signature","submit_replication":"https://pith.science/pith/HQQNQWHDORSYWL2EIP6YT2DDEP/action/replication_record"}},"created_at":"2026-07-05T11:49:22.776978+00:00","updated_at":"2026-07-05T11:49:22.776978+00:00"}