0 references
Multi-index Retrieve and Rerank with Sequence-to-Sequence Model
Annals of Computer Science and Information Systems2023Vol. 35, pp. 1277–1279
Abstract
This paper presents the solution to PolEval 2022 Task 3: Passage Retrieval.The main goal of the task, was to retrieve relevant text passages for the query.There were three different domains of passages: wikipedia passages, allegro faq and legal documents.The proposed solution incorporated both dense and lexical indexes, as well as, reranking model and reached 67.44 NDCG@10 score in official evaluation.
Related Papers
- → General literary index: Index of subjects.*(1865)
- → Business and economic books 1876–1983: subject index, author index, title index(1984)
- → Author index to volumes 18/19 (1979), subject index, materials index(1979)
- → Back Matter (“Bibliography”, “General Index”, “Index of Manuscripts and Archival Sources”, Index of Personal Names and Titles”, “Index of Place Names”, “Index of Scriptoria, Manuscript Groups, Scribes, and Illuminators”, “Index of Texts”)(2021)
- → Back Matter (“Bibliography”, "Index of Anaphoras", "Index of Manuscripts", "Index of Biblical Citations", "Index of Ancient Authors and Writings", "Index of Places and Regions", "General Index")(2023)