View source for 20240037190. MULTI-OUTPUT HEADED ENSEMBLES FOR PRODUCT CLASSIFICATION simplified abstract (RAKUTEN GROUP, INC.)

=MULTI-OUTPUT HEADED ENSEMBLES FOR PRODUCT CLASSIFICATION=


==Organization Name==

[[:Category:RAKUTEN GROUP, INC.|RAKUTEN GROUP, INC.]]


[[Category:RAKUTEN GROUP, INC.]]

==Inventor(s)==

[[:Category:Hotaka Shiokawa of Tokyo (JP)|Hotaka Shiokawa of Tokyo (JP)]][[Category:Hotaka Shiokawa of Tokyo (JP)]]

[[:Category:Pradipto Das of Tokyo (JP)|Pradipto Das of Tokyo (JP)]][[Category:Pradipto Das of Tokyo (JP)]]

==MULTI-OUTPUT HEADED ENSEMBLES FOR PRODUCT CLASSIFICATION - A simplified explanation of the abstract==


This abstract first appeared for US patent application 20240037190 titled 'MULTI-OUTPUT HEADED ENSEMBLES FOR PRODUCT CLASSIFICATION

==Simplified Explanation==

The patent application describes a method and system for classifying items using multi-output headed ensembles. The method involves receiving one or more text input sequences and tokenizing them into first tokens within estimator threads. The method also includes outputting item classifications based on the output of the estimator threads. Additionally, the method involves applying a backpropagation algorithm to update network weights, defining optimal network parameters using cross-validation, and mapping the first tokens to an embedding space within the estimator threads.

* The method involves receiving text input sequences and tokenizing them into first tokens within estimator threads.
* Item classifications are generated based on the output of the estimator threads.
* A backpropagation algorithm is applied to update network weights connecting neural layers in the estimator threads.
* Optimal network parameters are defined using cross-validation with respect to the estimator threads.
* The first tokens are mapped to an embedding space within the estimator threads.

Potential Applications:
* Text classification: This technology can be used to classify text items such as documents, articles, or social media posts into different categories or topics.
* Sentiment analysis: The method can be applied to analyze the sentiment or emotion expressed in text input sequences, allowing for sentiment classification or opinion mining.
* Spam detection: By classifying text input sequences, this technology can be used to identify and filter out spam or unwanted messages.

Problems Solved:
* Efficient classification: The method provides a way to classify items accurately and efficiently by utilizing multi-output headed ensembles and optimal network parameters.
* Text understanding: By tokenizing and mapping text input sequences, the method helps in understanding the content and context of the text, enabling better classification.

Benefits:
* Improved accuracy: The use of multi-output headed ensembles and optimal network parameters enhances the accuracy of item classification.
* Scalability: The method can handle multiple text input sequences and efficiently classify them, making it scalable for large-scale applications.
* Flexibility: The system can be adapted to different domains or languages by training the estimator threads with relevant data.


==Original Abstract Submitted==

an item classification method and system using multi-output headed ensembles, that can include receiving one or more text input sequences at one or more first estimator threads corresponding to the one or more text input sequences. the method can also include tokenizing the one or more text input sequences into one or more first tokens within the one or more first estimator threads. in addition, the method can include outputting one or more item classifications based on an output of the one or more first estimator threads. further, the method may include applying a backpropagation algorithm to update network weights connecting neural layers in the first estimator threads, defining an optimal setting of network parameters using cross-validation with respect to the first estimator threads, and mapping the one or more first tokens to an embedding space within the one or more first estimator threads.



[[Category:G06K9/62]]
[[Category:G06N3/08]]