What is Name Normalization and meta data aggregation?

You can apply custom aggregation rules to meta data for assignees, inventors and IPCs to have them considered as one during the similarity calculation and clustering process.

Setting up Name Normalization

Create an aggregation rule file in .xlsx, .tsv, or .txt format
Upload the file under More… → Name Normalization
Select the aggregation rule on Options tab before starting the analysis

Creating an Aggregation Rule

Matching Methods

MATCHING METHOD IDENTIFIER
Perfect match full
Partial match partial
Forward match forward
Backward match backward

Create an Excel file starting with the name you want to appear in the analysis
In the next column, enter the matching method
In the following columns enter the text to be used for aggregation
Start a new row to enter multiple rules in one file

If using .tsv or .txt, follow the same steps and separate each section with a tab

Format

Aggregated word 1[TAB]matching method identifier[TAB]aggregation target[TAB]aggregation target…

Applicant: abc[TAB]partial[TAB]abc corporation[TAB]ABC co
Inventor: John Smith[TAB]full[TAB]JOHN SMITH[TAB]John SMITH[TAB]john smith
IPC: G06N[TAB]forward[TAB]G06N 5

For example: BOEING Company[TAB]partial[TAB]Boeing[TAB]BOEING[TAB]The Boeing. This will aggregate names such as Boeing, Inc., BOEING Inc, The Boeing Corporation USA, etc. in the dataset into BOEING Company.