Issues with facet values


We use faceted filters in our index. Among values we sometimes have similar words like “hélice” and “helice”. When facet values are built it looks like Algolia “merges” these two values in a single one. Does anyone knows about this behavior, and if it is possible to configure so that words with accented letters are not considered as duplicates ?

Thanks in advance for your help.

Hi Nicolas,

The Algolia indexing engine removes diacritics as part of the data normalization process. You can tell the engine not to do this for particular diacritics using the keepDiacriticsOnCharacters parameter.

You can read more about the data normalization process and this setting here:

Hi Chuck,

We made the changes accordingly, it works like a charm !
Thank you so much for your answer.

Kind regards,

1 Like

Great news! If you can mark this as a “solution” it will help future users.