• Product: CDI in Primo / PrimoVE

In CDI, a new search method has been introduced.
As opposed to PC, where each search was identified and performed in a singular language, in CDI, the search is being performed in multiple languages, regardless of the UI language.
This search is being sent to the different CDI servers (where each server is handling the search in different language) and the rules of the languages are being applied to the search.

For example:
1. The English server is providing the same results for Decolonizing / Decolonising since this conversion appear in the English Synonyms file.
2. The English server is providing the same results for 'Obstetrics and Gynecology' and 'Obstetrics & Gynecology' since Ampersand is mapped to 'and' in English.
Since 'and' is a stop word in English, it is being removed.

However, rest of the servers, get the searched query and might return relevant results for the specific language.
For 1st example, Decolonizing will return X results from Spanish server, while Decolonising will return Y records from Spanish server.
For 2nd example, 'and' is kept since it is not a Spanish (for example) stop word, therefore the server will provide for 'Obstetrics and Gynecology' X results and Y for 'Obstetrics & Gynecology'.

Currently, the &\and mapping apply in all fields except the author field. 

This can cause a result counter difference between the searches since the records returned from the non-English servers are different for the different searches.

This behavior is by design and was implemented in order to provide as many as relevant results, from multiple languages.

• Article last edited: 13-Apr-2021