18480166. DEEP LEARNING FOR MULTIMEDIA CLASSIFICATION simplified abstract (Samsung Electronics Co., Ltd.)
Contents
- 1 DEEP LEARNING FOR MULTIMEDIA CLASSIFICATION
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 DEEP LEARNING FOR MULTIMEDIA CLASSIFICATION - A simplified explanation of the abstract
- 1.4 Simplified Explanation
- 1.5 Potential Applications
- 1.6 Problems Solved
- 1.7 Benefits
- 1.8 Potential Commercial Applications
- 1.9 Possible Prior Art
- 1.10 Unanswered Questions
- 1.11 Original Abstract Submitted
DEEP LEARNING FOR MULTIMEDIA CLASSIFICATION
Organization Name
Inventor(s)
Sunil Bharitkar of Stevenson Ranch CA (US)
DEEP LEARNING FOR MULTIMEDIA CLASSIFICATION - A simplified explanation of the abstract
This abstract first appeared for US patent application 18480166 titled 'DEEP LEARNING FOR MULTIMEDIA CLASSIFICATION
Simplified Explanation
The computer-implemented method described in the abstract involves using text information from a media content item's title and a trainable model to improve accuracy in classifying the item. The trainable model utilizes text to numeric-vector embeddings in a sequence for classification, optimizing word embedding model parameters or latent semantic analysis dimensions using the text information, and a classifier model to maximize accuracy.
- Utilizing text information from a media content item's title and a trainable model for classification improvement
- Using text to numeric-vector embeddings in a sequence for classification
- Jointly optimizing word embedding model parameters or latent semantic analysis dimensions
- Employing a classifier model to maximize accuracy in classifying media content items
Potential Applications
This technology could be applied in various industries such as media and entertainment, e-commerce, and information retrieval systems for more accurate content classification.
Problems Solved
This technology solves the problem of inaccurate or inefficient classification of media content items, leading to better organization and retrieval of information.
Benefits
The benefits of this technology include improved accuracy in classifying media content items, enhanced search and recommendation systems, and more efficient data organization.
Potential Commercial Applications
The potential commercial applications of this technology could include content recommendation systems, targeted advertising platforms, and information retrieval tools.
Possible Prior Art
One possible prior art could be the use of machine learning models for text classification in various industries, but the specific optimization techniques described in this patent application may be novel.
Unanswered Questions
How does this technology handle multi-language content classification?
The abstract does not mention how the trainable model deals with media content items in multiple languages. It would be interesting to know if the system can accurately classify content in different languages.
What is the computational overhead of implementing this technology?
The abstract does not provide information on the computational resources required to utilize this technology. Understanding the computational overhead could be crucial for organizations considering implementing this system.
Original Abstract Submitted
One embodiment provides a computer-implemented method that includes utilizing text information obtained from a title of a media content item and a trainable model for improving accuracy for classification of the media content item. The trainable model is utilized using a sequence of text to numeric-vector embeddings for classification of the media content item. At least one of a word embedding model parameter or a latent semantic analysis dimension is jointly optimized using the text information, and a classifier model for maximizing accuracy of the classification of the media content item.
