Skip to content

Semantic search

Semantic search uses an understanding of the meaning of the search phrase to match products with similar semantic content. This approach allows for flexibility in matching, as products do not have to contain every word in the query.

Info

Please note that semantic search is enabled only for searches resulting in zero results.

Setting up

This feature requires BAIA

If BAIA is not already enabled in your environment please reach out to your sales representative.

Semantic search can be enabled and disabled using the EnableSemanticSearch flag on the General settings endpoint API. Using the API, it is possible to enable the experimental feature intent-focused semantic search feature. Please contact your solution partner for help to set this up.

If a Unified Search returns zero results, a fallback search is performed using semantic search along with the Did You Mean suggestion.

Use cases

Semantic search significantly enhances the user experience by understanding the intent and contextual meaning behind search queries. This advanced search capability is particularly beneficial in two primary scenarios.

Synonym recognition

Semantic search inherently understands the relationships between words, including synonyms, without the need for manual entry in the search system.

Example

Consider a scenario where a user searches for coat. Traditional searches might only return products explicitly labeled with the word coat However, semantic search recognizes jacket as a related term, particularly highlighting long jackets that closely resemble coats in style and function.

Complex queries

Semantic search excels at interpreting long-winded or complex queries that may not have exact matches in the product database.

Example

A user enters a detailed query such as outdoor pool my grandkids can have in A traditional search might struggle to find products that match all the specific terms such as outdoor, pool and grandkids. Semantic search, on the other hand, can parse the query to understand the core intent - to find an outdoor pool that children will enjoy.

Data usage

Semantic search works by analyzing the semantic meaning embedded in product descriptions. To ensure an accurate and comprehensive semantic representation for search purposes, all attributes marked as searchable are used. Attributes not marked as searchable do not contribute to the semantic search process.

Example

In a bookstore with searchable attributes like Author and Genre, a search for Science fiction by Isaac Asimov will return books by Asimov in the science fiction genre, as these attributes directly match the search terms. Books with mentions of Asimov or science fiction in non-searchable sections won't appear in the results.

Improving results

Semantic search results can be improved by increasing the quality and specificity of product data. Semantic search is based on understanding the intent and contextual meaning behind a search query, rather than simply matching keywords

Warning

Semantic search doesn't use dictionaries, so semantic search results can't be improved by changing synonyms, hyponyms, etc.

Enhance product descriptions

Write detailed and clear product descriptions that accurately reflect the product's features, benefits, and use cases. Ensure that the language used is aligned with the terms customers might use to search for these products.

Instead of embedding the users search phrase directly BAIA will take a qualified guess from context on what the user is searching for and then use this to improve the search results. This is what is called intent focused semantic search. This does not affect search speed but do increase the number of used tokens.

For instance is the user is search for kornisjons but BAIA have been told through the search context that the site is food and grocery retailer it will search for the semantic meaning of Cornichons Pickled Cucumbers this will make it easier to find the correct products.

Search context

An important part of intent-focused semantic search is to provide BAIA with a good description of the context the user searches in. This is done during setup using the API. This will heavily improve the results. A good context description is just a short per segment description of the shop the search solution is used on for example:

my-example-shop.dk is large danish e-commerce site selling products for DIY projects, such as knitting sewing drawing and wood work.