Algolia’s search engine does not currently support intra-word matching, which means it does not identify matches that occur within words. For example, a search for "book" would not return results that contain "rebooking".
This limitation affects how Algolia performs with languages that do not use spaces between words, such as Japanese, Thai, or Chinese.
Why this affects search behavior
Algolia relies on tokenization, which typically splits text into individual words based on spaces or punctuation. In languages like Thai or Japanese, where words are written without spaces, Algolia cannot correctly identify or segment individual terms.
As a result, a search query using a word from these languages may return no matches, even if the content exists in the index.
Workaround
To enable effective search in these languages, the input content would need to be pre-processed or segmented with spaces between words before being indexed. This allows Algolia to treat each term as a separate searchable token.
For improved support in these use cases, external word segmentation tools can be used during indexing to break the text into space-separated tokens.