Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Introduction of a novel word embedding approach based on technology labels extracted from patent data

Mark Standke, Abdullah Kiwan, Annalena Lange, Silvan Berg

TL;DR

A word embedding approach using statistical analysis of human labeled data to produce accurate and language independent word vectors for technical terms is introduced.

Abstract

Diversity in patent language is growing and makes finding synonyms for conducting patent searches more and more challenging. In addition to that, most approaches for dealing with diverse patent language are based on manual search and human intuition. In this paper, a word embedding approach using statistical analysis of human labeled data to produce accurate and language independent word vectors for technical terms is introduced. This paper focuses on the explanation of the idea behind the statistical analysis and shows first qualitative results. The resulting algorithm is a development of the former EQMania UG (eqmania.com) and can be tested under eqalice.com until April 2021.

Introduction of a novel word embedding approach based on technology labels extracted from patent data

TL;DR

A word embedding approach using statistical analysis of human labeled data to produce accurate and language independent word vectors for technical terms is introduced.

Abstract

Diversity in patent language is growing and makes finding synonyms for conducting patent searches more and more challenging. In addition to that, most approaches for dealing with diverse patent language are based on manual search and human intuition. In this paper, a word embedding approach using statistical analysis of human labeled data to produce accurate and language independent word vectors for technical terms is introduced. This paper focuses on the explanation of the idea behind the statistical analysis and shows first qualitative results. The resulting algorithm is a development of the former EQMania UG (eqmania.com) and can be tested under eqalice.com until April 2021.

Paper Structure

This paper contains 14 sections, 1 equation, 4 figures, 1 table.

Table of Contents

  1. 1. Introduction
  2. 2. State of the art
  3. 3. EQMania approach
  4. 4. Results
  5. 5. Outlook and Conclusion
  6. Acknowledgements

Figures (4)

  • Figure 1: Meta Data types relevant for EQMania approach. The IPC/CPC classification will be used to train the presented algorithm.
  • Figure 2: Rake extracting key phrases from patents, for each key phrase four vectors, one for each meta data type, are generated. The vector with the orange frame contains information about the technological area and forms the word embedding base. The other three vectors contain information about the inventor, institution and time.
  • Figure 3: EQMania Word similarities displayed in 2D. The words form clusters according to the different technology areas they belong to.
  • Figure 4: Exemplary results of the search result as displayed in the user interface. The left shows most similar key phrases, while the right shows usage of key phrase over time.