Taxonomy Bot: Are they for real?

Taxonomy Bot: Are they for real? IP professionals often segment patents into different technology categories of interest to have a better understanding of the evolving technology trends and competitors’ activity. Generally, a comprehensive taxonomy is prepared by an expert and then manual categorization is done by going through the patent details. This process is manageable when there are up to 2000 Patents. It becomes a really hard and time-consuming task if technology is really broad and it has more than 2000 patents.

Automatic Classification & Challenges

The methods for automatic classification have been around for some time; however, the adaptability of such classification systems is really low due to their low explainability and the black box approach of AI. When your job depends upon the explainability and validation of why a result was considered or not considered while evaluating a patent, technology, or article, explainability is one of the most important concerns.

Our Approach

XLSCOUT Taxonomy Bot brings together man and machine. 

This works for you in a 4-step process:

• Set a detailed custom taxonomy by combining the corpus, machine suggestions, and your inputs.
• Apply it to the “patent set” you want to categorize.
• The categorization and graphical dashboard are ready in a few seconds.
• Additional option to train the taxonomy with seed and anti-seed patents, which have verified an accuracy of 95%.

The benefits of using XLSCOUT Taxonomy Bot are:

• Easy to Use
• Fast
• No Data limits
• Accurate
• Reduce Costs
• Explainable
• Repeatable
• Reinforced learning