Medical NLP Vector Dimensionality Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current natural language processing techniques for medical sentences require high-performance information processing apparatuses, increasing costs and failing to accurately consider the order of words, leading to misclassification of medical sentences with different meanings as similar.
Innovation Solution
An information processing apparatus and method that acquire a token sequence and context information vector from medical records, transforming them into a high-dimensional feature vector for more accurate output sequence generation, balancing calculation cost and processing capability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-performance information processing apparatuses such as GPU are used to accurately analyze medical sentences, then the processing capability and accuracy are improved, but the calculation cost and burden on medical practice increase
Solution Approach 1:
The patent transforms the input vector (combining token sequence and context information) into a high-dimensional feature vector through a connection weight matrix. This dimensional transformation enables the system to capture complex relationships and achieve accurate medical sentence analysis without requiring high-performance GPU apparatuses, thus resolving the contradiction between analysis accuracy and calculation cost
2Ease of operation
If conventional natural language processing techniques are used that do not consider word order, then the processing simplicity is maintained, but the accuracy of distinguishing medical sentences with different meanings deteriorates
Solution Approach 1:
The patent segments the medical sentence into individual tokens and maintains their sequential order, creating a token sequence that preserves word order information. This segmented approach allows the system to accurately distinguish between medical sentences with different meanings while keeping the processing method straightforward and computationally efficient
Solution Approach 2:
By transforming the segmented token sequence into a high-dimensional feature space through matrix multiplication, the patent enables the system to capture complex semantic relationships and word order patterns without complicating the basic processing framework, thus maintaining ease of operation while improving accuracy
3Measurement precision
If the dimension of feature vector is increased to improve processing capability, then the analysis accuracy is improved, but the calculation complexity increases
Solution Approach 1:
The patent achieves high-dimensional feature representation through a single matrix multiplication operation using a connection weight matrix. This approach transforms the input vector into a high-dimensional feature vector that captures complex relationships, improving analysis accuracy while maintaining calculation simplicity through a straightforward linear transformation rather than complex multi-layer processing
Data Source
AI summary
In order to attain an object to provide a technique in which calculation cost and a processing capability are well balanced and which is applicable to natural language processing in medical practice, an information processing apparatus includes: an acquisition means (21) for acquiring a token sequence obtained from a medical sentence in an electronic medical record and a context information vector obtained from context information of the electronic medical record; and an output sequence generation means (22) for carrying out an output sequence generation process for generating an output sequence from the token sequence and the context information vector, the output sequence generation process including a process for transformation into a high-dimensional feature vector that has a higher dimension than a sum of a dimension of the token sequence and a dimension of the context information vector.