Pre-training and/or transfer learning for sequence taggers
US9875736B2 · kind B2 · utility
15Cited by
3References
16Claims
0Family size
Assignee
Inventors
Key dates
| Filing date | Feb 19, 2015 |
| Grant date | Jan 23, 2018 |
| Priority date | — |
| Expiry date | Oct 20, 2035 |
Classification
- Technology area (CPC G)Physics
- CPC primaryG10L2015/0631
- WIPO fieldComputer technology
- WIPO sectorElectrical engineering
Abstract
Systems and methods for pre-training a sequence tagger with unlabeled data, such as a hidden layered conditional random field model are provided. Additionally, systems and methods for transfer learning are provided. Accordingly, the systems and methods build more accurate, more reliable, and/or more efficient sequence taggers than previously utilized sequence taggers that are not pre-trained with unlabeled data and/or that are not capable of transfer learning/training.
Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.