18244543. MULTIMODAL SENTIMENT CLASSIFICATION simplified abstract (Snap Inc.)

MULTIMODAL SENTIMENT CLASSIFICATION

Organization Name

Snap Inc.

Inventor(s)

Jianfei Yu of Los Angeles CA (US)

Luis Carlos Dos Santos Marujo of Culver City CA (US)

Venkata Satya Pradeep Karuturi of Marina del Rey CA (US)

Leonardo Ribas Machado das Neves of Marina Del Rey CA (US)

Ning Xu of Irvine CA (US)

William Brendel of Los Angeles CA (US)

MULTIMODAL SENTIMENT CLASSIFICATION - A simplified explanation of the abstract

This abstract first appeared for US patent application 18244543 titled 'MULTIMODAL SENTIMENT CLASSIFICATION

Simplified Explanation

The patent application describes a neural network for sentiment classification in social media posts that takes into account different entities and includes image data.

• The neural network includes separate subnetworks for the left, right, and target entities mentioned in a social media post.
• An image network generates representation data from images associated with the post.
• The output of the left, right, and target entity subnetworks is combined and weighted with the representation data from the image network.
• The combined data is used to classify the sentiment of the entity mentioned in the post.

Potential Applications

• Social media sentiment analysis
• Brand monitoring and reputation management
• Customer feedback analysis
• Market research and consumer insights

Problems Solved

• Accurate sentiment classification in social media posts
• Handling multiple entities mentioned in a post
• Incorporating image data for sentiment analysis

Benefits

• Improved understanding of sentiment towards specific entities in social media
• More accurate analysis of customer opinions and feedback
• Enhanced brand monitoring and reputation management capabilities
• Better insights for market research and decision-making

Original Abstract Submitted

Sentiment classification can be implemented by an entity-level multimodal sentiment classification neural network. The neural network can include left, right, and target entity subnetworks. The neural network can further include an image network that generates representation data that is combined and weighted with data output by the left, right, and target entity subnetworks to output a sentiment classification for an entity included in a network post.