Label matching

Search for entities that might represent the same real-world object based on their identities and labels and on specified criteria. For example, you can compare letters only, digits only, or choose to ignore characters such as spaces or tabs.

Label matching is specialized form of entity matching that does not consider criteria such as attribute instances and database properties.

If the entity has an identity, it is used when searching for a match. If the entity does not have an identity, for example if the entity is charted from a data source, the label is used. To override this behavior, select the Use All Identifiers check box in the Match Entities area of the Setup Label Matching window. All identities and labels are compared.

To prevent matches based on identity and label alone, you can match only entities of the same type or of the same type and with a specific semantic type assigned. For example, matching "Sam STEELE (Male)" with "Sam STEELE (Female)" based only upon their identity and label would not be a good match.

Entity Matching works by comparing a set of characters between one entity and another. You can choose to include only letters, only digits, ignore certain characters, and ignore case. For example, one data source might represent telephone numbers as three groups of digits that are separated by a hyphen (123-456-789). Another data source might represent telephone numbers without separation but with an extension number (123456 ext 789). The digits are the same in both cases. You can use label matching to compare digits only, ignoring the hyphens to match the entities and merge them into a single entity on your chart.

Label matching reports the search results as sets of potential matches, which are known as matched sets.