The most time-consuming and labor-intensive stage of patent analysis is, without a doubt, technology classification. WIPS is proud to announce the global launch of PRISM, our upgraded AI patent classification solution. PRISM is a deep-learning-based tool designed to smartly automate repetitive and complex classification tasks, simultaneously boosting both speed and accuracy.
Specifically engineered for the global state, PRISM supports four major languages-English, Chinese (simplified), Japanese and Traditional Chinese. This allows for consistent classification standards even in multi-lingual environments, significantly expanding the scope and efficiency of international IP analysis.
 |
| https://gl.wipsprism.com |
The Challenge: Why is Patent Analysis So Difficult?
Patent
analysis goes far beyond simple data organization; it is the backbone of
R&D strategy, IP portfolio management, market/competitor analysis, and
litigation risk management.
In
practice, experts must subdivide vast amounts of patent data into specific
technology, industry, and product units. This classification phase is
notoriously grueling, often accounting for 50% to 60% of the total analysis workload.
The Role of PRISM: The Core of Automated Classification
PRISM
was designed to solve this exact bottleneck. As an AI-driven solution, it can
be applied across the entire core analysis pipeline: from screening valid data
and classifying technologies to data refining and statistical analysis.
Technical Excellence of
PRISM
At
the heart of PRISM lies a Deep Learning
training model. While traditional classification methods suffer from
high labor costs, long turnaround times, and inconsistent results, PRISM
overcomes these limits. By combining multiple deep learning models with
preliminary classifiers and applying noise-reduction techniques, PRISM ensures
a faster, more consistent technical classification process.
Two Strategic AI
Classification Modes
PRISM
offers two distinct classification methods tailored to your specific analysis
goals:
1. Supervised Learning AI Classification When a user uploads
previously classified patent data, PRISM learns the patterns to create an
optimized classification model and noise filter. It then automatically
classifies new patent populations based on this learned logic.
2. Unsupervised learning AI clustering even without prior training data,
Users can upload a
patent population and the AI will automatically form clusters based on
technical similarity.
With a
hierarchical structure (up to 4 depths), this mode is ideal for exploring new
technologies or initial technology mapping.
Key Features & Visualization
PRISM
provides AI training/classification, matrix analysis, and clustering functions.
It visualizes classification results into an X-Y Axis Matrix, enabling intuitive technology positioning and
competitive landscape analysis.
A New Standard for IP
Intelligence
PRISM
is a solution that elevates patent analysis by automating the most resource-heavy
stage of the process. Through deep learning and its dual-mode
(Supervised/Unsupervised) approach, it supports everything from routine
repetitive analysis to the discovery of emerging technologies.
WIPS PRISM transforms patent classification from a mere
"pre-processing task" into a strategic decision-making asset.
Shift your IP workflow to a faster, smarter standard
with PRISM.
Stay tuned! In our next issue (#38), we will dive deeper into the
specific features of PRISM Global.
Inquiries: For detailed product information, please contact us at global@wips.co.kr
