Home List your patent My account Help Support us

Method to identify the document structure (sections, references)

[Category : - SOFTWARES]
[Viewed 453 times]

An inventive approach for identifying the structure of unstructured documents, whether they are digital native or scanned, regardless of the language they are written in (examples of documents in Chinese and Hebrew are depicted in the illustrations). This method, known as ITS (Identify The Structure), employs a heuristic approach that outperforms AI-based methods in terms of speed and accuracy.

In a continuation-in-part, the patent also encompasses the extraction of data from documents and the organization of that data into hierarchies based on the document structure.

The method offers several practical use cases, including:
1. Transforming unstructured documents into structured and navigable formats and converting internal references into clickable links.
2. Extracting data from a large collection of documents using Deep Learning techniques by applying preprocessing methods to optimize performance and output quality.
3. Enabling real-time translation of lengthy documents by presenting them section by section, which proves particularly useful compared to translating the entire document at once.
4. Improving the efficiency of LLM-based assistants (e.g., GPT assistants) that interact with lengthy documents by structuring the text, parallelizing computations, and adding annotations.

For more information and practical demonstrations of the applications, please visit Link

Patent publications:

Asking price:

Make an offer

Rate this patent

Great invention
Viewed: 453 times

Seller:

sixtyfourer (Italy)

Seller's website

Ask a question
or start a negotiation

Contact the seller

Share on social media