Rule-Based Extraction and Entity Recognition
Using rule-based extraction and enrichment, it is possible to easily and with a large amount of flexibility, extract different entities based on your own defined rules.
The rule-based extraction is available at the very heart of the Mindbreeze semantic pipeline – meaning that extracted entities can either be used to enrich the search experience or be used in combination with our other services to perform more complex tasks such as lookups or synthesize metadata.
Adding rule-based extraction is very simple within the configuration settings of your Mindbreeze InSpire management center. Pattern extraction rules in the form of regular expressions can be defined, and entities can be extracted from anywhere within the indexed document or from a specific location. The exact format of the extracted entity can also be defined.
Here is a quick video tutorial for a clearer and visual representation of what this looks like:
Curious to learn more about rule-based extraction and entity recognition? We have two previous blog posts that may help!
Rule-based or AI-based: The technologies behind chatbots