NLP Solutions for Patent Services 

People who deal with patent procedures usually need to go above and beyond in the analysis and comparisons of multiple documents. Patent search services are a confusing and demanding task. And that’s where one can bring automatization using Natural Language Processing, Machine Learning algorithms with top-quality engineering. 

In this article, we present which tasks of patent analytics and research can be solved using Intellexer NLP services. 

We bring automatization to these tasks using the following NLP processes:

• Comparison of patents on the same subject matter;

• Quick detection of plagiarism or copyright infringement.

A demo version for this solution is Intellexer Comparator (http://demo.intellexer.com/document_comparator_demo). Intellexer Comparator is a semantic solution specially developed to eliminate the hassles of content comparison. With its help, it is possible to accurately compare documents content and set the degree of similarity between them.

Intellexer Comparator analyzes each text document and converts it into a specific form that represents the essential meaning of the document. In this new representation, concrete words, phrases, and semantic relations between them, initially used as the termbase for the documents, acquire a generalized structure and meaning. As a result, the processing of information takes place at the level of the possible meanings of each word and at the level of the ideas that each sentence and the context, in general, may express.

The degree of proximity between documents is calculated within the range of 0-100%, where «0» means "absolutely different texts" and «100» means "the same text".

Concepts & Relations Extraction                       

• ​​​Statistical analysis of patents content

• Technology/Industry trends analysis

• Monitoring of competitors activity 

• Creating Patent Landscape (see figure)

Concepts are mainly noun groups that we extract from texts. Concept Extraction is available on our demo http://demo.intellexer.com/ (see tab “Clusterizer”).

Firstly, Semantic Search is conducted using Synonyms and Paraphrases. Then, concepts are extracted from the found Patents. Selection of Similar Patents candidates using Concepts Comparison is performed. And the final stage is the Deep Analysis of Similar Patents using Patent Index.

Using this algorithm the following tasks can be solved: Prior Art Search, Patent Novelty Analysis, Patent Infringement Analysis.

Patent Indexing                       

• Smart database creation. Using Machine Learning together with the rule-based approach Intellexer extracts main concepts from claims, and finds the most relevant 

• Patent Index is a sophisticated semantic profile of a patent that is built with Machine Learning algorithms and expert rules. See example below

We embrace a combined approach in the solutions presented above, incorporating statistics and rules with Machine Learning Frameworks. This approach works well for analyzing different various documents in a specific field, such as patents.

How can you contact us?

Feel free to contact us at quotes@intellexer.com. 

December 8, 2021

Get Started

Intellexer Summarizer

Application based on Intellexer API that performs:

• Document summarization
• Concept mining
• Entity extraction
• Summary rearrangement according to the selected items

Get access Online

Summarizer Network Edition

Download x86 (32 bit)

Download x64 (64 bit)

Note: After installation use your private API key to register the software. Number of requests per month and maximum size of the documents depend on your API key permissions.