US 11,941,363 B2
Information processing method and apparatus, and storage medium
Fan Dong Meng, Shenzhen (CN); Yun Long Liang, Shenzhen (CN); Jin Chao Zhang, Shenzhen (CN); Jie Zhou, Shenzhen (CN); and Jin An Xu, Shenzhen (CN)
Assigned to TENCENT TECHNOLOGY (SHENZHEN) COMPANY LTD, Shenzhen (CN)
Filed by TENCENT TECHNOLOGY (SHENZHEN) COMPANY LTD, Guangdong (CN)
Filed on Jun. 8, 2021, as Appl. No. 17/341,614.
Application 17/341,614 is a continuation of application No. PCT/CN2020/091733, filed on May 22, 2020.
Claims priority of application No. 201910452828.3 (CN), filed on May 28, 2019.
Prior Publication US 2021/0312135 A1, Oct. 7, 2021
Int. Cl. G06F 40/30 (2020.01); G06N 3/02 (2006.01)
CPC G06F 40/30 (2020.01) [G06N 3/02 (2013.01)] 20 Claims
OG exemplary drawing
 
1. An information processing method, performed by at least one processor of an information processing apparatus, comprising:
with respect to source data, encoding sub-data in the source data based on a target word feature vector to obtain hidden feature vectors corresponding to the sub-data, the target word feature vector representing a sentiment feature standard;
obtaining a word feature vector corresponding to the source data based on the hidden feature vectors corresponding to the sub-data; and
inputting the word feature vector into a preset sentiment classification network to obtain a result of sentiment polarity prediction of the source data,
wherein the sub-data in the source data comprises a plurality of pieces of the sub-data, and the encoding the sub-data comprises, except for a piece of the sub-data that is encoded first:
encoding each piece of the sub-data based on the target word feature vector, a preset gated nonlinear transformation model, and a hidden feature vector, which is obtained by encoding a previous piece of sub-data, the preset gated nonlinear transformation model being configured to perform nonlinear transformation on a corresponding piece of the sub-data to select a sub-feature vector that meets the target word feature vector.