Advanced Patent Monitoring using Artificial Intelligence and a Single-Source Database

Dr. Matthias Pötzl is Managing Director of Dennemeyer Octimine. Dennemeyer Octimine uses Natural Language Processing (NLP) methods, machine learning and artificial intelligence algorithms to analyze and compare millions of scientific and technical text documents (e.g. patents, scientific publications, scientific news,…) in seconds to retrieve the relevant information much faster than common methods. Founded in 2015 as a spin-off from the University of Munich (LMU) and the Max-Planck-Institute, the company is now part of the Dennemeyer IP Group.

The main product octimine is a semantic patent search engine (SaaS) that makes patent search easy and fast. There is no need for specialized knowledge about Boolean search operators or technology classes to perform prior art or freedom to operate (FTO) searches.
100 to 500 documents—that's the number of patents a technology company has to analyze each week to stay up-to-date with the patenting activity of the competition. Due to the annually increasing number of worldwide patent applications and the growing complexity of technologies, it is becoming increasingly difficult to keep up with the latest state of the art. Traditional approaches to analysis are based on creating a perfect Boolean search and running that search every week on the newly published patent documents. For the reasons mentioned above, this is becoming more and more difficult and the results of a conventional competition monitoring are often not very precise, resulting in the aforementioned 100-500 weekly documents.
There are, however, two factors that may improve the situation: Firstly, there are more and more data providers such as IFI CLAIMS, who offer access to a consolidated, worldwide patent database. In this area, IFI stands out for its uncomplicated access. The new machine translations of non-English documents allow for an almost worldwide analysis of patent documents in English. However, just as important as good data is the correct use of machine learning. In combination with keyword-based analysis, machine learning can contribute to a massive reduction of the analysis effort.
Due to further developments in the field of artificial intelligence, it is now possible for neural networks to understand the context of words in a patent text. Furthermore, such neural networks are able to learn different synonyms and writing styles, as well as to identify hidden patterns in the data. Since an individual neural network is created for each user profile, results are tailor-made for that user.

The latest developments from Dennemeyer Octimine show what this means for the end user. The company has been active in the field of AI and patent data for over 7 years ( Due to its proximity to the scientific research of the Max Planck Institute, the University of Munich and the Technical University of Munich, Octimine is known for a high standard of development. It does not use off-the-shelf algorithms, but is explicitly trained on patent data. 

Users can now use artificial intelligence in addition to classic, keyword-based monitoring to:
  1. Supplement existing analyses with the hits of the machine search and prioritize the analysis of hundreds of documents.
  2. Create new monitoring profiles within minutes. Existing technology portfolios can easily be used as training data for a new neural network. Just a few clicks and the profile is ready.
The ability to create customized profiles is especially helpful in making analysis more manageable. In most cases, around 30 documents per technology portfolio are identified per week using this method. Compared to 100-500 documents, the time (and cost) savings are obvious.

In early 2021, Dennemeyer Octimine will launch a completely new and extremely powerful patent monitoring software. This will enable newbies to set up an innovative and efficient monitoring system in just a few steps and will give experienced users the opportunity to add new methods to existing systems quickly and easily.

In summary, the increasing availability of high-quality data and the ever-advancing AI algorithms will make IP management much easier. High-quality analyses can be performed in less time and with less effort.