Need for full text documents

Due to the sheer amount of information (about 150 million documents just in our database), most of the patent information or the non-patent literature is available in the form of abstracts, thus making possible some quick browsing. Using this abbreviated information (title, abstract, author), it is possible to perform efficient patent or non-patent search on a global basis with the goal of narrowing down search results to a manageable relevant quantity of documents.

If, however, after reading the abstract information, a patent seems to be relevant, one has to diligently study the full text of this document or article; i.e. the complete description and the claims within the patents, because these are the places, where the key information of a patent is unveiled (“the inventive core”).

Full-text availability

In ChemAnalyser, in following sections of the underlying databases full-text documents are available:

  • Patent Literature

    • For currently 31 of the 103 Intellectual Property Offices, including the most important ones like CN, JP, US, KR, and WO, patent full text documents are available; for an exact list see here.

  • Non-Patent Literature

    • PubMedCentral (PMC) contains a large collection of biomedical full text articles, mainly in English language
    • Selected „Web“ Sources with special focus on the domains of chemistry, drugs, nutrition, and health also offer full text contributions of different kinds in different languages.

Many key players publish patents in Non-English languages

In a large number of cases one will find that the relevant patent full text is published in a non-English language, e.g. Chinese, Japanese or Korean, to just name a few of the most important ones. So there is a large potential for important information to be overlooked in the patent retrieval process.

Machine translations help avoiding this blind spot

To substantially reduce this dangerous blind spot, our patent database is augmented by the English machine translations of the full texts for more than 20 of the most important patent offices, e.g. CN, JP, and KR. These translations are based on specially developed dictionaries and thesauri for different technological domains; they include intelligent translation engines and are specifically adjusted to Intellectual Property specific wording and nomenclature.   

You can browse through these texts individually for better understanding; in addition, however, these translations are also included in our patent searching algorithms, both semantically and field-based, and they are also chemically annotated.

The conclusion is, machine translations help you achieving the best possible and most extensive search results.

Human translations for highest accuracy

If in these two afore-mentioned steps a patent is identified which seems to be important for a future business decision or a legal case, higher translation accuracy is needed. In such a case a professional human translation will be necessary. 

Our database is equipped with a one-click feature, automatically requesting a human translation directly from the machine translated text. By clicking this feature, professional linguists perform these critical translations without involving machine translation support. The translation will be managed in the shortest possible time.