System and method for enhancing search precision with multi-level passage ranking
View Patent ↗A method is presented for enhancing search results by segmenting documents into smaller passages and utilizing those passages as the search unit. This method integrates the passage rankings from two search models to produce a new document ranking and arranges the documents accordingly. The ranking of document is also rearranged based on the proportion of passages taken from the same document versus the total number of passages in that document. The final search ranking system combines document-level and passage-level search rankings to rank documents. This method ensures that when conducting a passage search, the search results consider the general relevance of the entire document, which leads to better performance compared to searching only by document-level or passage-level searches.
1 . A computer-implemented method for providing a user with search results corresponding to a query entered by the user from a passage corpus including a plurality of passages extracted from each document of a document corpus, the method comprising:
(a) extracting and arranging, by a first retrieval model, from the passage corpus N passages in correspondence with the query, the N passages being globally retrieved across the passage corpus for the query;
(b) re-ranking, by a second search model, the N passages based on the query;
(c) generating an integrated ranking of the N passages by integrating the results in step (a) and the re-ranking in step (b) for the N passages;
(d) arranging M documents containing the N passages with the integrated ranking for said N passages, wherein M is less than or equal to N;
(e) arranging the M documents based on a relationship between the number of passages extracted from a particular document among the N passages and the total number of passages in the particular document; and
(f) determining a final ranking for the M documents by integrating the results in step (d) and the results in step (e), by rank-fusing the results, the rank-fusing being performed even when at least one of the rankings lacks a score for a given document;
wherein, in step (e), the M documents are arranged either in an order from documents with a smaller value given by the relation (Np−np)/Np to documents with a larger value, or in an order from documents with a larger value given by the relation np/Np to documents with a smaller value, where np denotes a count of passages of the given document among said N passages globally retrieved in step (a), and Np denotes a total number of passages in the given document.
2 . The method of claim 1 , wherein the integration of the results in step (c) and step (f) is performed by Reciprocal Rank Fusion (RRF).
3 . An apparatus for providing a user with search results corresponding to a query entered by the user from a passage corpus comprising a plurality of passages extracted from each document of a document corpus, comprising:
at least one processor; and
at least one memory for storing computer-executable instructions,
wherein the computer-executable instructions stored in the at least one memory make the at least one processor to perform the following steps:
(a) extracting and arranging, by a first retrieval model, from the passage corpus N passages in correspondence with the query, the N passages being globally retrieved across the passage corpus for the query;
(b) re-ranking, by a second search model, the N passages based on the query;
(c) generating an integrated ranking of the N passages by integrating the results in step (a) and the re-ranking in step (b) for the N passages;
(d) arranging M documents containing the N passages with the integrated ranking for said N passages, wherein M is less than or equal to N;
(e) arranging the M documents based on a relationship between the number of passages extracted from a particular document among the N passages and the total number of passages in the particular document; and
(f) determining a final ranking for the M documents by integrating the results in step (d) and the results in step (e), by rank-fusing the results, the rank-fusing being performed even when at least one of the rankings lacks a score for a given document;
wherein, in step (e), the M documents are arranged either in an order from documents with a smaller value given by the relation (Np−np)/Np to documents with a larger value, or in an order from documents with a larger value given by the relation np/Np to documents with a smaller value, where np denotes a count of passages of the given document among said N passages globally retrieved in step (a), and Np denotes a total number of passages in the given document.
4 . The apparatus of claim 3 ,
wherein at least part of the documents in the document corpus have titles and each passage of the passage corpus includes a title of a document of which said each passage is included as a part.
5 . The apparatus of claim 3 , wherein integration of the results in step (c) and step (f) is performed by a reciprocal rank fusion (RRF).
6 . The apparatus of claim 3 , wherein at least one of the first search model and the second search model is an artificial intelligence based search model.