Sentiment Polarity Prediction via Sub-Data Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current deep CNN network architectures for sentiment analysis often overlook content sentiment analysis, leading to inaccurate sentiment polarity predictions due to the omission of local feature content during feature extraction.
Innovation Solution
An information processing method that encodes sub-data in source data using a target word feature vector to obtain hidden feature vectors, which are then used to generate a word feature vector for input into a sentiment classification network for accurate sentiment polarity prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If deep CNN network architecture extracts local features of window section for sentiment analysis, then the model can process fine-grained sentiment analysis work, but content sentiment analysis of local features is omitted leading to inaccurate sentiment polarity prediction
Solution Approach 1:
The patent segments the sentiment analysis process into multiple encoding stages: first encoding sub-data (local features) to obtain sub-feature vectors, then encoding aspect data to obtain aspect feature vectors, and finally combining them. This segmentation allows each stage to focus on specific features without losing content sentiment information, resolving the contradiction between fine-grained processing and content analysis completeness.
Solution Approach 2:
The patent performs preliminary encoding of sub-data using a target word feature vector before final sentiment polarity prediction. This preliminary action extracts and preserves content sentiment features from local features early in the process, ensuring they are not lost during subsequent processing stages and improving overall prediction accuracy.
2Device complexity
If local features and preset aspect features are used for data encoding without content sentiment analysis, then the processing flow remains simple, but the sentiment polarity representation becomes inaccurate
Solution Approach 1:
The patent introduces an intermediary encoding process that uses a target word feature vector as a mediator between sub-data and the final sentiment prediction. This intermediary step transforms local features into meaningful sub_feature vectors that preserve content sentiment information, improving accuracy without significantly increasing overall system complexity.
Solution Approach 2:
The patent changes the encoding parameters by introducing a target word feature vector and performing multiple encoding operations with different parameters. The first encoding uses the target word feature vector to capture content sentiment, while subsequent encodings combine multiple feature types, achieving high accuracy through parameter variation rather than structural complexity.
Data Source
AI summary
Embodiments of the disclosure provide an information processing method, an information processing apparatus, and a storage medium. The method includes: obtaining 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. According to the embodiments of the disclosure, the accuracy of sentiment polarity prediction may be improved.