Invention Title:

SYSTEMS AND METHODS FOR MODIFICATION OF MACHINE LEARNING MODEL-GENERATED TEXT AND IMAGES BASED ON USER QUERIES AND PROFILES

Publication number:

US20260311235

Publication date:
Section:

Physics

Class:

G06Q30/0627

Inventors:

Assignee:

Applicant:

Smart overview of the Invention

A novel system is designed to generate personalized text and image outputs based on user queries and profiles. It employs machine learning models to tailor content that aligns with user-specific interests and preferences. By incorporating user profiles, the system enhances the relevance of generated content, allowing for more accurate and engaging results. This approach addresses limitations in existing content generation systems that often fail to account for individual user preferences or rely solely on pre-existing data.

Background

Artificial intelligence, particularly large language models (LLMs), has significantly advanced content creation. These models predict language tokens to generate text, while image generators create visuals from user prompts. Despite their capabilities, these models can be limited by biases in training data, leading to inaccurate or irrelevant outputs. Additionally, they may suffer from hallucinations, producing content that does not reflect reality. This invention seeks to overcome these challenges by providing user-tailored content.

Personalization

The system enhances personalization by using user profiles to generate content that reflects individual preferences. For example, if a user frequently searches for computing systems with specific features, the system will highlight these attributes in generated results, even if they are not explicitly mentioned in the current query. This personalized approach reduces inaccuracies and improves user trust by aligning outputs with user expectations and previous interactions.

Functionality

The system processes a user's query by retrieving their profile, which includes past queries and preferences. It then obtains descriptions of items relevant to the query and profile. Using a machine learning model, it generates personalized text and images that emphasize features important to the user. This process ensures that the generated content is both relevant and engaging, improving the user's experience and satisfaction with the search results.

Application

In practical terms, the system can handle queries for specific items, such as computing systems with desired hardware components. By integrating user profiles, it produces outputs that are not only accurate but also tailored to the user's history and preferences. This personalized content generation reduces the likelihood of inaccuracies and hallucinations, providing users with realistic and relevant results that enhance engagement and trust in the system.