#community-help

Resolving Issues with Synonyms and Typo Tolerance in a Document

TLDR Slavik was having trouble with synonyms and typo tolerance while searching for documents. Jason suggested using one-way vs multi-way synonyms and configuring typo-tolerance parameters.

4mo

Join the chat

May 15, 2023 (4 months ago)

Photo of md5-ca6495d5be926db80e09aabf066f4b8b

Slavik

04:10 PM

Hello!
May I ask you for your help, please?

I have a document whose name field value is Volkswagen e-Golf 35 kWh (2017 - 2020).
I added the synonyms to Volkswagen (screenshot 1) as vw, VW, vewe. But when I am trying to search by vw or VW - there are no documents :(

cURL:

curl '' \
  -H 'authority: ' \
  ...
  --data-raw '{"searches":[{"q":"vw","query_by":"name","sort_by":"name:asc","highlight_full_fields":"name","collection":"X","page":1,"per_page":8}]}' \
  --compressed

I am using the alias for my index and searching inside it.
The name field structure is: { name: 'name', type: 'string', facet: false, sort: true }

$Image 1 for Hello! May I ask you for your help, please? I have a document whose name field value is <code>Volkswagen e-Golf 35 kWh (2017 - 2020)</code>. I added the synonyms to <code>Volkswagen</code> (screenshot 1) as <code>vw, VW, vewe</code>. But when I am trying to search by <code>vw</code> or <code>VW</code> - there are no documents :( cURL: <pre><code>curl '<https://xxx.typesense.net/multi_search?x-typesense-api-key=API-KEY>' \ -H 'authority: <http://xxx.typesense.net|xxx.typesense.net>' \ ... --data-raw '{"searches":[{"q":"vw","query_by":"name","sort_by":"name:asc","highlight_full_fields":"name","collection":"X","page":1,"per_page":8}]}' \ --compressed</code></pre> I am using the alias for my index and searching inside it. The <code>name</code> field structure is: <code>{ name: 'name', type: 'string', facet: false, sort: true }</code>$

Photo of md5-8813087cccc512313602b6d9f9ece19f

Jason

04:16 PM

With your current settings in the screenshot, where you’ve set Volkswagen as the root, only searching for Volkswagen will return results with that word, and also the words you’ve mentioned under Synonyms

04:16

Jason

04:16 PM

Here’s a description of what one-way vs multi-way synonyms mean (first paragraph): https://typesense.org/docs/0.24.1/api/synonyms.html#synonyms

Slavik

04:17 PM

Yes. Sorry, found an issue.
Adding vw as a root and Volkswagen as a synonym resolved an issue.

Jason

04:19 PM

Side note: Typesense is case in-sensitive, so you don’t have to add vw and VW, just lower case is sufficient

Slavik

04:22 PM

Thank you!
also, another question.

I have a document whose name field value is BMW 225xe (2019 - 2022).

When I am trying to search by bmw -> it works fine.
When I am trying to search by bwm -> nothing. The same for bm .

04:23

Slavik

04:23 PM

  --data-raw '{"searches":[{"q":"Bwm","query_by":"name","sort_by":"name:asc","highlight_full_fields":"name","collection":"x","page":1,"per_page":8}]}' \
  --compressed

04:24

Slavik

04:24 PM

Is it possible to resolve on a query side on improving Typo-Tolerance parameters or only by synonyms?

Jason

04:28 PM

By default typo-tolerance only kicks-in when a search term is more 4 characters, but you can configure this using min_len_1typo and min_len_2typo

https://typesense.org/docs/0.24.1/api/search.html#typo-tolerance-parameters

Slavik

04:29 PM

Thank you, Jason!

Typesense

Lightning-fast, open source search engine for everyone | Knowledge Base powered by Struct.AI

Indexed 2764 threads (79% resolved)

Join Our Community Manage subscription

Similar Threads

Phrase Search Relevancy and Weights Fix

Jan reported an issue with phrase search relevancy using Typesense Instantsearch Adapter. The problem occurred when searching phrases with double quotes. The team identified the issue to be related to weights and implemented a fix, improving the search results.

111

6mo

Utilizing Vector Search and Word Embeddings for Comprehensive Search in Typesense

Bill sought clarification on using vector search with multiple word embeddings in Typesense and using them instead of OpenAI's embedding. Kishore Nallan and Jason informed him that their development version 0.25 supports open source embedding models. They also resolved Bill's concerns regarding search performance, language support, and limitations in the search parameters.

225

2mo

Typesense Search Solution Issues

Rolando faced incorrect search results using Typesense. Kishore Nallan suggested changing typo parameters and upgrading Typesense version. However, undesired results persisted and need further investigation.

26mo

Querying and Indexing Multiple Elements Issues

Krish queried fields with multiple elements, which Kishore Nallan suggested checking `drop_tokens_threshold`. Krish wished to force OR mode for token, but Kishore Nallan admitted the feature was missing. Krish was able to resolve the issue with url encoding.

9mo

Fine-Tuning Typesense Queries for Better Results with Plural Words

Sergio seeks advice on improving Typesense queries for plural words. Kishore Nallan suggests increasing the typo tokens threshold and adjusting the min_len_1typo and min_len_2typo params. Sergio confirms better results after making these changes.

4mo