Patent Literature

Why to search in patent literature?

At the beginning of every research or development project, a thorough study of the state of the art should be conducted. This is particularly advisable for three reasons:

  • Gaining insight into the latest progress in the respective field.
  • Avoiding duplication of effort.
  • Protecting the project against possible future infringement proceedings as soon as possible.

One of the most effective information sources for studying the state of the art of a technology is the patent literature:

  • One estimates that  80% of current technical knowledge can only be found in patent documents.
  • Patents – due to their strict publishing procedures – contain very detailed technical information which often cannot be found in any other publicly availabe information.
  • Information in patents is always up-to-date, since the patent documents in their majority have to be publicly published 18 months after the first filing at the latest.
  • Commercially available databases (e.g. Sem-IP.com) contain the complete set of patents worldwide and offer tools to efficiently search in huge collections, including semantic algorithms and statistical support.

Worldwide coverage

The Sem-IP.com database which is at the foundation of ChemAnalyser, currently contains records of about 100M patents worldwide, i.e. from 103 patent issuing authorities of all continents.

To give you a quick impression of this worldwide coverage, the approximate filing numbers of just the most important authorities (patent offices) in the last 20 years are listed below:

Chemical patents

In an ever increasing amount, chemical drawings are added to the text for illustration of the respective structure; standardized file formats for this purpose, implemented in our database, are CDX and MOL (information about the atoms and chemical bonding of a molecule).

To extract the chemistry specific knowledge as efficiently as possible, ChemAnalyser applies proprietary automated knowledge extraction processes based on chemical named entity recognition (NER) in a semantic context. Furthermore, ChemAnalyser recognizes and extracts relevant relationships between the NERs or identifies correlations between the chemical structure of a chemical compound and its biomedical impact.