US 12,462,154 B2
Method and system for aspect-level sentiment classification by merging graphs
Xiaochen Hou, Mountain View, CA (US); Peng Qi, Mountain View, CA (US); Guangtao Wang, Cupertino, CA (US); Zhitao Ying, Palo Alto, CA (US); Jing Huang, Mountain View, WA (US); Xiaodong He, Beijing (CN); and Bowen Zhou, Beijing (CN)
Assigned to CHINABANK PAYMENT (BEIJING) TECHNOLOGY CO., LTD., Beijing (CN)
Filed by Chinabank Payment (Beijing) Technology Co., Ltd., Beijing (CN)
Filed on Feb. 21, 2022, as Appl. No. 17/676,775.
Prior Publication US 2023/0267322 A1, Aug. 24, 2023
Int. Cl. G06N 3/08 (2023.01); G06F 18/20 (2023.01); G06F 18/213 (2023.01); G06F 40/205 (2020.01); G06N 3/048 (2023.01)
CPC G06N 3/08 (2013.01) [G06F 18/213 (2023.01); G06F 18/29 (2023.01); G06F 40/205 (2020.01); G06N 3/048 (2023.01)] 20 Claims
OG exemplary drawing
 
1. A system comprising a computing device, the computing device comprising a processer and a storage device storing computer executable code, wherein the computer executable code, when executed at the processor, is configured to:
receive an aspect term-sentence pair, the aspect term-sentence pair comprising an aspect term and a sentence, and the sentence comprising the aspect term;
embed the aspect term-sentence pair to obtain an embedding of each word in the sentence;
parse the sentence using a plurality of parsers to obtain a plurality of dependency trees;
perform edge union on the plurality of dependency trees by taking a union of all edges in the plurality of dependency trees from the plurality of parsers to obtain a merged graph, wherein each node and each edge in the plurality of dependency trees are included in the merged graph, and wherein parent-to-child relations between nodes are retained in the merged graph while discarding parser-specific relation types;
represent each node in the merged graph by corresponding one of the embeddings of the words to obtain a relation graph;
perform a relation neural network on the relation graph to obtain updated relation neural network;
extract a hidden representation of the aspect term from the updated relation neural network to obtain an extracted representation of the aspect term; and
classify the aspect term based on the extracted representation to obtain a predicted classification label of the aspect term.