Skip to content

Fix/synonym tweaks #2229

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Nov 18, 2025
Merged

Fix/synonym tweaks #2229

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
f075e3b
Adjust analyzers and query retriever priorities to improve search res…
cotti Nov 17, 2025
ff28793
Add document headings to the queryable content
cotti Nov 17, 2025
d1f45f6
Add custom tokenizer and simplify analyzers
cotti Nov 17, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
25 changes: 17 additions & 8 deletions src/Elastic.Markdown/Exporters/Elasticsearch/ElasticsearchExporter.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -62,7 +62,6 @@ DistributedTransport transport
SearchInferenceId = endpoint.NoElasticInferenceService ? null : ".elser-2-elastic"
});


public abstract class ElasticsearchExporter<TChannelOptions, TChannel> : IDisposable
where TChannelOptions : CatalogIndexChannelOptionsBase<DocumentationDocument>
where TChannel : CatalogIndexChannel<DocumentationDocument, TChannelOptions>
Expand Down Expand Up @@ -149,14 +148,14 @@ protected static string CreateMappingSetting(string synonymSetName) =>
"analysis": {
"analyzer": {
"synonyms_analyzer": {
"tokenizer": "whitespace",
"tokenizer": "group_tokenizer",
"filter": [
"lowercase",
"synonyms_filter"
]
},
"highlight_analyzer": {
"tokenizer": "standard",
"tokenizer": "group_tokenizer",
"filter": [
"lowercase",
"english_stop"
Expand All @@ -176,7 +175,11 @@ protected static string CreateMappingSetting(string synonymSetName) =>
}
},
"tokenizer": {
"path_tokenizer": {
"group_tokenizer": {
"type": "char_group",
"tokenize_on_chars": [ "whitespace", ",", ";", "?", "!", "(", ")", "&", "'", "\"", "/", "[", "]", "{", "}" ]
},
"path_tokenizer": {
"type": "path_hierarchy",
"delimiter": "/"
}
Expand Down Expand Up @@ -243,17 +246,24 @@ protected static string CreateMapping(string? inferenceId) =>
},
"stripped_body": {
"type": "text",
"search_analyzer": "highlight_analyzer",
"search_analyzer": "synonyms_analyzer",
"term_vector": "with_positions_offsets"
}
},
"headings": {
"type": "text",
"search_analyzer": "synonyms_analyzer"
},
{{(!string.IsNullOrWhiteSpace(inferenceId) ? AbstractInferenceMapping(inferenceId) : AbstractMapping())}}
}
}
""";

private static string AbstractMapping() =>
"""
, "abstract": { "type": "text" }
, "abstract": {
"type": "text",
"search_analyzer": "synonyms_analyzer"
}
""";

private static string InferenceMapping(string inferenceId) =>
Expand All @@ -278,5 +288,4 @@ public void Dispose()

GC.SuppressFinalize(this);
}

}
1 change: 1 addition & 0 deletions src/api/Elastic.Documentation.Api.Core/Search/SearchUsecase.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -72,6 +72,7 @@ public record SearchResultItem
public required string Title { get; init; }
public required string Description { get; init; }
public required SearchResultItemParent[] Parents { get; init; }
public string[]? Headings { get; init; }
public float Score { get; init; }
public string? HighlightedBody { get; init; }
}
16 changes: 13 additions & 3 deletions src/api/Elastic.Documentation.Api.Infrastructure/Adapters/Search/ElasticsearchGateway.cs
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,6 +37,9 @@ internal sealed record DocumentDto
[JsonPropertyName("url_segment_count")]
public int UrlSegmentCount { get; init; }

[JsonPropertyName("headings")]
public string[] Headings { get; init; } = [];

[JsonPropertyName("parents")]
public ParentDocumentDto[] Parents { get; init; } = [];

Expand Down Expand Up @@ -88,10 +91,15 @@ public ElasticsearchGateway(ElasticsearchOptions elasticsearchOptions, ILogger<E

var lexicalSearchRetriever =
((Query)new PrefixQuery(Infer.Field<DocumentDto>(f => f.Title.Suffix("keyword")), searchQuery) { Boost = 10.0f, CaseInsensitive = true }
|| new MatchPhrasePrefixQuery(Infer.Field<DocumentDto>(f => f.Title), searchQuery) { Boost = 9.0f }
Copy link
Member

@Mpdreamz Mpdreamz Nov 18, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This is for a follow up but jotting thoughts down here, we probably do not want to have all these prefix queries score so high.

I think we should set up an explicit completion fields:

For general purpose: https://www.elastic.co/docs/reference/elasticsearch/mapping-reference/search-as-you-type

And explicit .prefix multifields on title and maybe headers using (edge)NGram tokenizer (or use the one provided by the search_as_you_type field).

|| new MatchQuery(Infer.Field<DocumentDto>(f => f.Title), searchQuery) { Operator = Operator.And, Boost = 8.0f }
|| new MatchBoolPrefixQuery(Infer.Field<DocumentDto>(f => f.Title), searchQuery) { Boost = 6.0f }
|| new MatchQuery(Infer.Field<DocumentDto>(f => f.Abstract), searchQuery) { Boost = 4.0f }
|| new MatchQuery(Infer.Field<DocumentDto>(f => f.StrippedBody), searchQuery) { Boost = 3.0f }
|| new MatchQuery(Infer.Field<DocumentDto>(f => f.Abstract), searchQuery) { Operator = Operator.And, Boost = 5.0f }
|| new MatchQuery(Infer.Field<DocumentDto>(f => f.StrippedBody), searchQuery) { Operator = Operator.And, Boost = 4.5f }
|| new MatchQuery(Infer.Field<DocumentDto>(f => f.Headings), searchQuery) { Operator = Operator.And, Boost = 4.5f }
|| new MatchQuery(Infer.Field<DocumentDto>(f => f.Abstract), searchQuery) { Operator = Operator.Or, Boost = 4.0f }
|| new MatchQuery(Infer.Field<DocumentDto>(f => f.StrippedBody), searchQuery) { Operator = Operator.Or, Boost = 3.0f }
|| new MatchQuery(Infer.Field<DocumentDto>(f => f.Headings), searchQuery) { Operator = Operator.Or, Boost = 3.0f }
|| new MatchQuery(Infer.Field<DocumentDto>(f => f.Parents.First().Title), searchQuery) { Boost = 2.0f }
|| new MatchQuery(Infer.Field<DocumentDto>(f => f.Title), searchQuery) { Fuzziness = 1, Boost = 1.0f }
)
Expand Down Expand Up @@ -129,7 +137,8 @@ public ElasticsearchGateway(ElasticsearchOptions elasticsearchOptions, ILogger<E
e => e.Title,
e => e.Url,
e => e.Description,
e => e.Parents
e => e.Parents,
e => e.Headings
)
)
)
Expand Down Expand Up @@ -193,6 +202,7 @@ private static (int TotalHits, List<SearchResultItem> Results) ProcessSearchResp
Url = doc.Url,
Title = doc.Title,
Description = doc.Description ?? string.Empty,
Headings = doc.Headings,
Parents = doc.Parents.Select(parent => new SearchResultItemParent
{
Title = parent.Title,
Expand Down
Loading