18094245. PREDICTING TOPICS OF POTENTIAL RELEVANCE BASED ON RETRIEVED/CREATED DIGITAL MEDIA FILES simplified abstract (Google LLC)
PREDICTING TOPICS OF POTENTIAL RELEVANCE BASED ON RETRIEVED/CREATED DIGITAL MEDIA FILES
Organization Name
Inventor(s)
Robert Rose of Boulder CO (US)
Qun Cao of Mountain View CA (US)
PREDICTING TOPICS OF POTENTIAL RELEVANCE BASED ON RETRIEVED/CREATED DIGITAL MEDIA FILES - A simplified explanation of the abstract
This abstract first appeared for US patent application 18094245 titled 'PREDICTING TOPICS OF POTENTIAL RELEVANCE BASED ON RETRIEVED/CREATED DIGITAL MEDIA FILES
Simplified Explanation
Implementations described in this patent application leverage digital media files retrieved and/or created by users to predict and determine topics of potential relevance to the users.
- Users' digital media files are used as input for trained machine learning models to generate output indicating objects detected in the files.
- The actual digital media files are not provided to a remote computing system, only data indicative of the detected objects.
- Information associated with the detected objects can be retrieved and proactively output to the user.
- The frequency of objects occurring across a corpus of digital media files is considered when determining the relevance of a detected object to a user.
Potential Applications
This technology has potential applications in various fields, including:
- Content recommendation systems: The predicted topics of relevance can be used to recommend relevant content to users, such as articles, videos, or products.
- Personalized advertising: The detected objects can be used to deliver targeted advertisements to users based on their interests and preferences.
- Social media platforms: The technology can be used to enhance user experience by providing relevant content and improving content discovery.
- E-commerce platforms: The predicted topics can be used to personalize product recommendations and improve search results for users.
Problems Solved
This technology addresses several problems:
- Information overload: By predicting and determining topics of potential relevance, users can more easily find and discover content that aligns with their interests.
- User engagement: By proactively outputting information associated with detected objects, users are more likely to engage with the content and stay engaged with the platform.
- Personalization: The technology enables personalized recommendations and advertisements, enhancing the user experience and increasing the likelihood of user satisfaction.
Benefits
The benefits of this technology include:
- Improved content discovery: Users can find relevant content more easily, saving time and effort.
- Enhanced user experience: Proactively providing information associated with detected objects increases user engagement and satisfaction.
- Personalization: By considering the user's own digital media files, the technology can provide highly personalized recommendations and advertisements.
- Privacy preservation: Only data indicative of the detected objects is provided to the remote computing system, ensuring the privacy of the user's digital media files.
Original Abstract Submitted
Implementations are described herein for leveraging digital media files retrieved and/or created by users to predict/determine topics of potential relevance to the users. In various implementations, digital media file(s) created and/or retrieved by a user with a client device may be applied as input across trained machine learning model(s), which in some cases are local to the client device, to generate output that indicates object(s) detected in the digital media file(s). Data indicative of the indicated object(s) may be provided to a remote computing system without providing the digital media file(s) themselves. In some implementations, information associated with the indicated object(s) may be retrieved and proactively output to the user. In some implementations, a frequency at which objects occur across a corpus of digital media files may be considered when determining a likelihood that a detected object is potentially relevant to a user.
