Rosette Name Indexer and Rosette Name Translator (RNI-RNT) provides the linguistic infrastructure and Java APIs to perform name matches, name searches, and translations across an expanding collection of languages and scripts.
You can use RNI-RNT to perform the following tasks:
For information about other Rosette products that can help with processing documents, extracting names, and additional text analytics, contact support@rosette.com.
Overview of Name Matching
The natural language processing algorithms employed by RNI use machine learning and cutting-edge NLP techniques to perform name matching. The match scores produced are a relative indication of how similar two names are, or a search name is to a name in an index; the higher the score the stronger the match. Customizations are available to tune and configure RNI to fit your business and data.
There are two common usage patterns in name and address matching: pairwise and index.
In pairwise matching, you have two names or addresses that you are comparing directly to one another. This comparison results in a single similarity score that indicates how similar the two names are.
With index matching, you have a single name or address that you are comparing to a list. This can be thought of as a search problem. You have a name and want to search are large list of records to find a match.
Index matching includes pairwise matching. When querying an index RNI performs a two-pass search:
Generate candidates: The first pass is designed to quickly generate a set of candidates for the second pass to consider.
Pairwise match: The query value is compared with each value returned by the first pass and a similarity score is calculated for each pair.