Patent · US Active

Transformer-based encoding incorporating metadata

US11893346B2 · kind B2 · utility

0Cited by
5References
19Claims
0Family size

Assignee

Inventors

Key dates

Filing dateMay 5, 2021
Grant dateFeb 6, 2024
Priority date
Expiry dateSep 30, 2041

Classification

  • Technology area (CPC G)Physics
  • CPC primaryG06V30/194
  • WIPO fieldComputer technology
  • WIPO sectorElectrical engineering

Abstract

From metadata of a corpus of natural language text documents, a relativity matrix is constructed, a row-column intersection in the relativity matrix corresponding to a relationship between two instances of a type of metadata. An encoder model is trained, generating a trained encoder model, to compute an embedding corresponding to a token of a natural language text document within the corpus and the relativity matrix, the encoder model comprising a first encoder layer, the first encoder layer comprising a token embedding portion, a relativity embedding portion, a token self-attention portion, a metadata self-attention portion, and a fusion portion, the training comprising adjusting a set of parameters of the encoder model.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.