Docket Alarm's New Entity Normalizer Endpoint Overhauls Search Precision and Recall

Searching for a company’s litigation record is rarely as simple as entering a name and clicking “search.” Companies may be listed on dockets under dozens of variations— parent corporations, subsidiaries, business units, misspellings and more. 

Missing relevant search results is not the only concern either. Searches for common business names may yield overly-inclusive results, including false-positives for “Apple Orchards Inc.” when intending to only locate results for “Apple” Computers.

Though the use of fuzzy searching can improve results, it never fully addresses the issue as it faces the same tradeoff between precision and recall as a well-crafted boolean search. 

Docket Alarm is excited to announce a solution— our latest AI endpoint, the Entity Normalizer. The Entity Normalizer solves normalization issues by delivering LLM-derived query language that is perfectly tailored to locate your targeted entity. It is easier than ever to run perfect client alerts and conduct comprehensive client research. 

The Entity Normalizer is simple to use. Send an API call to this endpoint with:

  1. the targeted entity;

  2. Whether the entity’s corporate group should be included; and

  3. A search limit value.

The Entity Normalizer returns an optimized query to be used for searching or tracking that entity. Now, the next time you search for “Apple” Computers, you can avoid retrieving results for “Apple Oil,” “Apple Valley Bank,” or even “Apple Grove Condominium Association.” Full information is available in our API Specification page.

Docket Alarm is an advanced search and litigation tracking service for the Patent Trial and Appeals Board (PTAB), the International Trade Commission (ITC), Bankruptcy Courts, and Federal Courts across the United States. Docket Alarm searches and tracks millions of dockets and documents for thousands of users.

view all posts