Automated Product Identification in Research Documents
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Researchers face challenges in accessing and identifying commercial products mentioned in scientific or research-related documents, such as antibodies and chemicals, due to the complexity of extracting relevant information from text.
Innovation Solution
A system comprising a server that receives input documents, divides them into tokens, computes scores for each token to determine if it corresponds to a commercial product, and provides output when the score exceeds a threshold, using machine learning to identify products by name, catalog number, manufacturer, and location, and displays information about identified products in a user-friendly format.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual extraction of product information from research documents is performed, then information accuracy can be maintained, but time consumption and labor intensity increase significantly
Solution Approach 1:
The patent replaces manual mechanical extraction processes with an automated computer-based system that uses machine learning models and natural language processing to identify and extract product information from research documents, thereby reducing time consumption while maintaining accuracy through algorithmic analysis
Solution Approach 2:
The system enables self-service extraction where the computer automatically processes documents, identifies product mentions, and retrieves product information without requiring manual intervention, allowing the system to serve itself in performing the extraction task
2Loss of information
If comprehensive product information is extracted from documents, then data completeness improves, but system complexity increases
Solution Approach 1:
The patent segments the extraction process into distinct modular components: a machine learning model for identifying product mentions, a natural language processing module for extracting attributes, and a database system for storing information. This segmentation allows comprehensive information extraction while managing complexity through modular architecture
Solution Approach 2:
The system employs a universal machine learning framework that can handle multiple product types (chemicals, biological reagents, equipment) and extract various attributes (name, catalog number, manufacturer, concentration) using the same core architecture, reducing overall system complexity through multi-functionality
3Productivity
If automated product identification is implemented, then processing speed increases, but identification accuracy may decrease
Solution Approach 1:
The patent implements feedback mechanisms where the machine learning model is trained on labeled data and continuously improved based on performance metrics. The system provides feedback loops for model retraining and optimization, ensuring high identification accuracy is maintained as processing speed increases through automated batch processing
Data Source
AI summary
Aspects of the present disclosure relate to identifying a product in a document. A server accesses a document including scientific or research-related text. The server divides the document into a plurality of tokens, each token comprising a part of the text that logically comprises a unit of information. The server computes, for each token in the plurality of tokens, a score corresponding to whether the token corresponds to a commercial product, the score being computed based on a list of features of commercial products and weights assigned to features in the list. The server determines that the score exceeds a threshold score. The server provides, in response to determining that the score exceeds the threshold score, an output representing that the token corresponds to the commercial product.