Traditional search tools use mathematical and statistical concepts to determine the closeness of a name match. Such an approach works well with minor spelling variations or typing errors, but cannot begin to address the broader variations in names caused by different transcription standards, or the phonetic similarities specific to each language, unless the parameters are set so widely that many non-relevant matches are also generated.
In recent years, some tools have adopted a halfway house approach, using mathematical algorithms augmented by dictionaries of name variations. While this clearly represents a step in the right direction, built in dictionaries can never accommodate variations of all possible names, and often do not deal well with the names of organisations and legal entities, which commonly comprise words not traditionally associated with identity data.
Organisations using such non-linguistic tools often have to compromise in order to make their hit handling processes viable and, in doing so, run the risk of missing real matches, with all the associated reputational, financial and regulary consequences that his can incur. However, with advances in computational linguistic knowledge it is no longer necessary to make this compromise, and high levels of assurance can be achieved without excessive investment in hit handling resources.
This need to balance a meticulous search with the operational requirement to minimise costly manual review is what makes Traphoty® the identity search solution for the business environment of the future.
