Bitext NAMER Cracks Named Entity Recognition

On November 19, the Beyond Search Web log published a brief analysis of our multilingual NER (Named Entity Recognition system) technology.

The post highlighted the challenges of handling Chinese personal names in English to enable accurate and consistent cross-tabulation for analysts, researchers, and investigators.

Similar issues arise with organizational names, such as “Sun City” (a place and enterprise) or aliases like “Yati New City” for “Shwe Koko”; and, in general, with any language that is written in non-Roman alphabet and needs transliteration.

In fact, these issues affect to all languages that do not use Roman alphabet including Hindi, Malayalam or Vietnamese, since transliteration is not a one-to-one function but a one-to-many and, as a result, it generates ambiguity the hinders the work of analysts.

With real-time data streaming into government software, resolving ambiguities in entity identification is crucial, particularly for investigations into activities like money laundering. The Bitext NAMER addresses these challenges, including:

1. Correctly and identifying generic names.

2. Assigning them a type: person, place, time, organization…

3. Resolving aliases, also known as (AKAs), and psuedonyms.

4. Distinguishing similar names linked to potentially unrelated entities (e.g., “Levo Chan”).

Bitext’s proprietary methods support more than 20 languages, with an additional 30 languages available on request.