The TurboPatent Patent Quality Report (PQR) leverages three key technologies
1. Natural Language Processing (NLP). The PQR NLP is based on the Apache Open NLP open source project, specially tuned and optimized to apply to the typical language and style of patent applications. NLP techniques are used throughout the analysis of the claim language and disclosure text to identify claim terms (noun phrases) and part names.
2. Big Data. The PQR grades all submitted content on a "curve." By analyzing a large, random sample of recently granted applications, it is able to present each quality metric individually, and a weighted summary score, in the context of a percentile rank within the sample population.
3. Machine Learning. The newest PQR metric, the Alice Risk, is based on a machine learning model trained on the outcomes of the examination process in tens of thousands of cases impacted by the Alice decision and subsequent changes in examination standards and practices.