US 12,450,432 B2
Method of entity recognition for data table, electronic device and storage medium
Yonghong Yu, Beijing (CN); Zhiyong Liu, Beijing (CN); Chao Xu, Beijing (CN); Yang Gao, Beijing (CN); and Xiaoming Zhao, Beijing (CN)
Assigned to BEIJING VOLCANO ENGINE TECHNOLOGY CO., LTD., Beijing (CN)
Filed by Beijing Volcano Engine Technology Co., Ltd., Beijing (CN)
Filed on Dec. 19, 2024, as Appl. No. 18/988,785.
Claims priority of application No. 202410030477.8 (CN), filed on Jan. 8, 2024.
Prior Publication US 2025/0225322 A1, Jul. 10, 2025
Int. Cl. G06F 40/30 (2020.01); G06F 16/334 (2025.01); G06F 40/174 (2020.01); G06F 40/186 (2020.01); G06F 40/279 (2020.01)
CPC G06F 40/279 (2020.01) [G06F 16/3344 (2019.01); G06F 40/174 (2020.01); G06F 40/186 (2020.01)] 18 Claims
OG exemplary drawing
 
1. A method of entity recognition for a data table, comprising:
obtaining a to-be-processed query statement described in a natural language;
encoding the to-be-processed query statement according to entity information of a to-be-query data table for each to-be-query data table in at least one to-be-query data table, to acquire at least one encoded result, wherein the at least one encoded result comprises an encoded result respectively corresponding to each to-be-query data table;
after encoding the to-be-processed query statement according to entity information of the to-be-query data table for each to-be-query data table in at least one to-be-query data table, to acquire at least one encoded result, respectively inputting the at least one encoded result into an entity recognition model, to acquire at least one entity prediction result, wherein the at least one entity prediction result comprises an entity prediction result respectively corresponding to each encoded result in the at least one encoded result, and the entity recognition model is configured to predict an entity included in input information according to the input information; and
determining a target entity prediction result corresponding to the to-be-processed query statement according to the at least one entity prediction result, wherein the target entity prediction result indicates the entity included in the to-be-processed query statement, and the target entity prediction result corresponds to a queried data table,
the encoding the to-be-processed query statement according to the entity information of the to-be-query data table comprises:
encoding a first word segmentation that is completely matched with an entity in the to-be-query data table in the to-be-processed query statement by using a first encoding manner; and/or
encoding a second word segmentation that is partially matched with the entity in the to-be-query data table in the to-be-processed query statement by using a second encoding manner, wherein the second encoding manner is different from the first encoding manner.