17947971. DEEP NAVIGATION VIA A MULTIMODAL VECTOR MODEL simplified abstract (eBay Inc.)
Contents
- 1 DEEP NAVIGATION VIA A MULTIMODAL VECTOR MODEL
- 1.1 Organization Name
- 1.2 Inventor(s)
- 1.3 DEEP NAVIGATION VIA A MULTIMODAL VECTOR MODEL - 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 Original Abstract Submitted
DEEP NAVIGATION VIA A MULTIMODAL VECTOR MODEL
Organization Name
Inventor(s)
Christopher Miller of San Jose CA (US)
Shubhangi Tandon of San Jose CA (US)
SenthilKumar Gopal of San Jose CA (US)
Selcuk Kopru of San Jose CA (US)
DEEP NAVIGATION VIA A MULTIMODAL VECTOR MODEL - A simplified explanation of the abstract
This abstract first appeared for US patent application 17947971 titled 'DEEP NAVIGATION VIA A MULTIMODAL VECTOR MODEL
Simplified Explanation
A multimodal embedding modifier enhances search results by generating a modified seed search selection embedding to improve user intent identification in online marketplaces.
- The multimodal embedding modifier allows users to navigate multiple modalities for an item.
- Users can select a search result and further modify it by inputting a modifier, such as a textual modifier.
- The modifier can be trained using various types of embeddings, including text, image, or a combination thereof.
Potential Applications
This technology could be applied in e-commerce platforms, online search engines, and recommendation systems to enhance user experience and improve search result accuracy.
Problems Solved
1. Improved user intent identification in online marketplaces. 2. Enhanced search result accuracy and relevance.
Benefits
1. Better user experience in navigating search results. 2. Increased efficiency in finding desired items. 3. Enhanced accuracy in identifying user intent.
Potential Commercial Applications
1. E-commerce platforms 2. Online search engines 3. Recommendation systems
Possible Prior Art
There are existing technologies in the field of natural language processing and image recognition that may have similarities to the multimodal embedding modifier, but further research is needed to identify specific prior art in this area.
=== What are the limitations of the multimodal embedding modifier in accurately identifying user intent? The abstract does not provide information on the potential limitations of the multimodal embedding modifier in accurately identifying user intent.
=== How does the training dataset impact the performance of the multimodal embedding modifier? The abstract does not mention how the training dataset, which includes various types of embeddings, impacts the performance of the multimodal embedding modifier.
Original Abstract Submitted
A multimodal embedding modifier generates a modified seed search selection embedding for providing a set of search results. The multimodal embedding modifier enhances the ability and accuracy of identifying a user's true intent when searching the online marketplace. For example, embodiments disclosed herein can allow a user to navigate multiple modalities for an item. In some embodiments, a user may select a search result corresponding to an initial search query, and further modify the selected search result by inputting a modifier (e.g., a textual modifier). The multimodal embedding modifier can be trained using a training dataset including a text embedding, an image embedding, another type of embedding, or a combination thereof.
