CNN Integrated Circuit for Ideogram Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current machine learning technologies face challenges in effectively processing natural languages using ideograms, as they struggle to combine and interpret the meaning of multiple ideograms within a single symbol, limiting their ability to understand complex linguistic structures.
Innovation Solution
A method utilizing a Cellular Neural Networks (CNN) based integrated circuit that creates a multi-layer two-dimensional symbol comprising a matrix of pixels, where each sub-matrix represents an ideogram, and employs a trained convolutional neural network model with bi-valued 3×3 filter kernels to learn the combined meaning of these ideograms, enabling efficient natural language processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If standard convolutional neural network models are used for processing multi-layer 2-D symbols containing multiple ideograms, then the model can process complex linguistic structures, but the computational complexity and processing time increase significantly
Solution Approach 1:
The patent segments the complex natural language processing task into multiple convolutional layers, where each layer processes specific features of the multi-layer 2-D symbol containing ideograms. This segmentation allows parallel processing of different linguistic features simultaneously, reducing overall processing time while maintaining the ability to handle complex structures.
Solution Approach 2:
The patent transforms the natural language input into a multi-layer two-dimensional symbol representation, adding spatial and hierarchical dimensions to the data structure. This dimensional transformation enables the CNN to process linguistic information through spatial convolutions, improving processing efficiency while capturing complex relationships between ideograms.
2Loss of information
If multiple ideograms are combined into a single multi-layer 2-D symbol for processing, then the representation of complex linguistic structures is improved, but the difficulty of interpretation and classification increases
Solution Approach 1:
The patent applies preliminary convolutional filtering operations to the multi-layer 2-D symbol before full classification. These preliminary convolutions extract and highlight key features of the combined ideograms, preparing the data for more efficient interpretation and reducing the difficulty of subsequent classification tasks.
Solution Approach 2:
The patent introduces convolutional filter kernels as intermediary elements that mediate between the complex multi-layer symbol and the classification output. These filters act as intermediaries that transform the complex ideogram combinations into simplified feature representations that are easier to interpret and classify.
3Productivity
If a CNN based integrated circuit is used for natural language processing, then the processing speed and efficiency are improved, but the device complexity increases
Solution Approach 1:
The patent merges multiple CNN processing functions into a single integrated circuit architecture. By combining the convolutional layers, pooling operations, and classification functions into one unified hardware structure, the system achieves high processing efficiency while managing device complexity through functional integration rather than separate components.
Data Source
AI summary
A string of natural language texts is received and formed a multi-layer 2-D symbol in a computing system. The 2-D symbol comprises a matrix of N×N pixels of K-bit data representing a “super-character”. The matrix is divided into M×M sub-matrices with each sub-matrix containing (N/M)×(N/M) pixels. K, N and M are positive integers, and N is preferably a multiple of M. Each sub-matrix represents one ideogram defined in an ideogram collection set. “Super-character” represents a meaning formed from a specific combination of a plurality of ideograms. The meaning of the “super-character” is learned by classifying the 2-D symbol via a trained convolutional neural networks model having bi-valued 3×3 filter kernels in a Cellular Neural Networks or Cellular Nonlinear Networks (CNN) based integrated circuit.