Once you have at least 5x examples each of both Positive and Negative patents in your Training Set, click 'Build Classifier'. 

This will introduce a new data column to the lists of patents you are reviewing. A Score, or number between 0.01 - 1.0, is assigned to each patent based on how relevant to the training set the machine believes it to be.

  • The higher the score, the more relevant the machine deems the patent to be. For example, a patent with a Score of 0.88 is considered to be more relevant to what you are interested in than a patent with a Score of 0.45. 
  • The column 'Classifier class' will populate to show whether the classifier thinks the patent is a Positive or Negative example of the technology you're looking for. 
  • To continue populating your Training Set, you will need to either confirm a patent as Positive or Negative, or reassign it. This will populate the 'User class' column.
  • The Score is a helpful indication of much training (examples of Positive and Negative) is still required for the machine to understand the parameters of your technology definition.

Note: If you mark something as “negative” which the classifier believes you have marked something similar as “positive” previously, it will continue to question you on this technology to ensure it’s understanding of what you are interested in is correct.

Did this answer your question?