Finding relationships in Trademark Data

Matt OverstreetJune 30, 2013

At the recent National Day of Civic hacking here at OSC we dug into a few ways to find relationships between Trademarks files with the USPTO.

If you’ve ever played with the US trademark data you’ll know that it’s both plentiful and scarce. There are lot’s of trademark fillings, each with the minimum possible data to make them uniquely identifiable.

That’s great for streamlined government and citizen anonymity, but no fun for finding the relationships between filings. We needed to suss out more information about the graph of trademarks.  That’s when we Eric and Wes tripped over the translations included in many of the patent filings.  We wondered if the term space for these translations might be smaller and more consistent then the space defined by the actual trademarks.  Translations were less likely to play games with spelling or grammar the way one might with the actual mark.

Some Hacking with the data and Neo4j resulted in an intriguing dataset that we are still unpacking.  Want to play with the data? Neo4J loaded with data is at this url:

rosetta_raw_queries rosetta_terms_related_to_cheese rosetta_terms_related_to_beer rosetta_concepts_related_to_hill_english_and_foreign rosetta_data_loaded_

More blog articles:

Let's do a project together!

We provide tailored search, discovery and analytics solutions using Solr and Elasticsearch. Learn more about our service offerings