{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,7]],"date-time":"2026-07-07T01:14:44Z","timestamp":1783386884609,"version":"3.54.6"},"reference-count":0,"publisher":"River Publishers","content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["JWE"],"abstract":"<jats:p>This study introduces task-injected layered hybrid retrieval-augmented generation (TILHR-RAG), a framework specifically designed for Persian to address the scarcity of native-language resources and the limitations of English-centric approaches. The architecture combines task-aware query augmentation, a layered retrieval strategy, and a hybrid semantic\u2013lexical retriever, all supported by a multi-stage pipeline that includes preprocessing, document chunking, question generation, and embedding. A novel mechanism for injecting task-specific vectors directs retrieval toward domain intent while preserving comparability across documents. The layered design operates in three stages: per-task frequently asked questions (FAQ) retrieval, hybrid document search using FAISS semantic similarity combined with BM25 keyword matching, and a fallback response generated by a large language model (LLM). This structure ensures both precision and robustness. Comprehensive experiments across five progressively refined configurations demonstrate that TILHR-RAG achieves the best balance among accuracy, efficiency, and scalability, reaching 89.67% semantic accuracy with moderate latency and memory consumption on NVIDIA A100 hardware. Further evaluations on low-resource graphics processing units (GPUs) confirm that accuracy remains stable under hardware constraints, although latency increases significantly. Moreover, multilingual E5 embedding models substantially improve retrieval and generation quality for Persian \u2013 outperforming ParsBERT and Sentence-BERT (SBERT) \u2013 by mitigating challenges such as orthographic variation and complex compound word structures. Taken together, these findings establish task-injected layered hybrid retrieval-augmented generation as a practical, reproducible, and resource-efficient blueprint for Persian question answering, advancing retrieval-augmented generation for low-resource languages without requiring costly large language model fine-tuning, while also offering adaptable strategies for broader multilingual applications.<\/jats:p>","DOI":"10.13052\/jwe1540-9589.2552","type":"journal-article","created":{"date-parts":[[2026,7,7]],"date-time":"2026-07-07T00:38:21Z","timestamp":1783384701000},"source":"Crossref","is-referenced-by-count":0,"title":["Hybrid Layered Retrieval and Task-aware Embeddings for Efficient Persian RAG Systems"],"prefix":"10.13052","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-1797-6910","authenticated-orcid":false,"given":"Toktam","family":"Zoughi","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ehsan","family":"Arianyan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Maryam","family":"Mahmoudi","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mahtab","family":"Aghdamifard","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Mohadese","family":"Nikoogoftar","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"5195","published-online":{"date-parts":[[2026,7,6]]},"container-title":["Journal of Web Engineering"],"original-title":[],"link":[{"URL":"https:\/\/journals.riverpublishers.com\/index.php\/JWE\/article\/download\/31201\/23998","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.riverpublishers.com\/index.php\/JWE\/article\/download\/31201\/23999","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/journals.riverpublishers.com\/index.php\/JWE\/article\/download\/31201\/23998","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,7,7]],"date-time":"2026-07-07T00:38:21Z","timestamp":1783384701000},"score":1,"resource":{"primary":{"URL":"https:\/\/journals.riverpublishers.com\/index.php\/JWE\/article\/view\/31201"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,7,6]]},"references-count":0,"URL":"https:\/\/doi.org\/10.13052\/jwe1540-9589.2552","relation":{},"ISSN":["1544-5976","1540-9589"],"issn-type":[{"value":"1544-5976","type":"electronic"},{"value":"1540-9589","type":"print"}],"subject":[],"published":{"date-parts":[[2026,7,6]]}}}