Emoji Sentiment Modeling for Electronic Response Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Determining likely responses to electronic communications, particularly those containing non-textual content such as emojis, is challenging due to the nuanced and complex nature of human interaction, which can lead to unintended negative reactions.
Innovation Solution
A computing device uses a first machine learning model to group emojis based on sentiment and a second model to select predicted responses, trained on historical message data, to improve the accuracy of response prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If non-textual content (emojis, images, audio) is included in electronic communications, then the expressiveness and information content of messages are improved, but the difficulty of determining likely responses increases
Solution Approach 1:
The patent segments the complex task of response prediction by separating textual content analysis from non-textual content (emoji) analysis. The system processes text and emojis independently through different machine learning models, then combines the results. This segmentation allows each component to be analyzed with appropriate methods, improving overall accuracy while managing complexity.
Solution Approach 2:
The patent introduces an intermediary processing layer that analyzes the relationship between textual content and emoji content. This intermediary layer detects potential sarcasm or negative sentiment that arises from the combination of text and emoji, which neither component alone would reveal. This mediator enables the system to capture nuanced meanings in multi-modal communications.
2Measurement precision
If sarcasm and hidden sentiment in messages are detected, then the accuracy of response prediction is improved, but the complexity of analysis increases
Solution Approach 1:
The patent applies preliminary action by training multiple specialized machine learning models in advance on historical data containing labeled sarcasm and sentiment examples. These pre-trained models can then quickly analyze new messages without requiring complex real-time computation. The system prepares response prediction capabilities beforehand, enabling accurate detection of sarcasm and hidden sentiment when messages are actually sent.
Solution Approach 2:
The patent changes parameters by using different machine learning models for different aspects of analysis (textual content, emoji content, combined analysis). Each model is optimized for specific parameters such as sentiment detection, sarcasm detection, or response likelihood prediction. This parameter-based approach allows the system to handle complexity by distributing analysis across multiple specialized models rather than one monolithic complex model.
Data Source
AI summary
Systems, methods, and apparatuses are described herein for performing sentiment analysis on electronic communications relating to one or more image-based communications methods, such as emoji. Message data may be received. The message data may correspond to a message that is intended to be sent but has not yet been sent to an application. Using a first machine learning model, one or more subsets of the plurality of emoji may be determined. The one or more subsets of the plurality of emoji may comprise one or more different types and quantities of emoji, and may each correspond to the same or a different sentiment. Using a second machine learning model, one or more emojis may be selected from the one or more subsets. The one or more emojis selected may correspond to responses to the message.