Skip to content

Commit 358c915

Browse files
RounakDhillon“Rounak
andauthored
Docs: Search Configuration Settings (#458)
Co-authored-by: “Rounak <“[email protected]”>
1 parent 6ed81b7 commit 358c915

File tree

8 files changed

+228
-0
lines changed

8 files changed

+228
-0
lines changed
Lines changed: 114 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,114 @@
1+
---
2+
title: Search Configuration Settings for Relevance, Ranking, and Field Boosting
3+
description: Detailed guide on customizing search relevance and behavior across a platform, including global rules, term boosting, field weighting, and asset-specific configurations.
4+
slug: /how-to-guides/admin-guide/search-configuration-settings
5+
---
6+
7+
# Search Configuration Settings
8+
9+
The **Search Configuration** section in OpenMetadata allows administrators to fine-tune the relevance and behavior of search results across the platform. You can define these settings at two levels:
10+
11+
- **Global Search Level**
12+
- **Data Asset Type Level** (e.g., Containers, Dashboards, APIs)
13+
14+
## 1. Global Search Configuration
15+
16+
These settings impact **all searches across data assets**. Use this to apply platform-wide rules for performance, security, and scoring.
17+
18+
### 🔗 Access Path:
19+
`Settings → Preferences → Search`
20+
21+
### Configurable Options:
22+
23+
| Setting | Description |
24+
|-----------------------------------|-----------------------------------------------------------------------------|
25+
| Enable Roles & Policies in Search | Toggle to include role-based access policies in search filtering. |
26+
| Max Aggregate Size | Maximum size of aggregated search results (default: `10000`). |
27+
| Max Result Hits | Maximum number of search hits returned in a single query (default: `10000`). |
28+
| Max Analyzed Offset | Controls how much of a field’s content is analyzed (default: `1000`). |
29+
30+
### Term Boost (Global)
31+
32+
Define boost values to give priority to specific field terms in the search scoring mechanism. For example, Tier levels:
33+
34+
| Field | Boost |
35+
|--------------|-------|
36+
| Tier.Tier1 | 50 |
37+
| Tier.Tier2 | 30 |
38+
| Tier.Tier3 | 10 |
39+
40+
- **Use Case:** Prioritize Tier 1 assets over lower-tiered ones in search results.
41+
42+
{% image
43+
src="/images/v1.10/how-to-guides/admin-guide/global.png"
44+
alt="Global Search Configuration"
45+
caption="Global Search Configuration"
46+
/%}
47+
48+
## 2. Data Asset Level Search Configuration
49+
50+
Search ranking can be further refined per **data asset type**, such as Containers, Dashboards, or APIs. These allow **domain-specific control** over what fields or values influence the ranking of search results.
51+
52+
### Matching Fields
53+
54+
Defines which fields contribute to determining relevance in search.
55+
56+
| Property | Description |
57+
|----------------|--------------------------------------------------|
58+
| Field | The metadata field (e.g., `name.keyword`) |
59+
| Match Type | Determines type of match (e.g., `Exact Match`) |
60+
| Weight | Numeric value defining the field’s importance |
61+
| Highlight Fields | Toggle to highlight this field in search preview |
62+
63+
**Example:**
64+
`name.keyword` set as **Exact Match** with **weight 20**
65+
66+
{% image
67+
src="/images/v1.10/how-to-guides/admin-guide/data-asset-level-search.png"
68+
alt="Data Asset Level Search Configuration"
69+
caption="Data Asset Level Search Configuration"
70+
/%}
71+
72+
73+
### Term Boost
74+
75+
Boosts specific field values to rank some records higher.
76+
77+
| Field | Boost |
78+
|--------------------|---------------|
79+
| dbtTags.model_tag_one | (Custom value) |
80+
81+
82+
83+
### Field Value Boost
84+
85+
Boost based on **numerical or date-based field values**, using conditions and modifiers.
86+
87+
**Configuration Fields:**
88+
89+
- **Field**: Choose the field to evaluate
90+
- **Impact**: Boosting factor
91+
- **Modifier**: *(none, log, sqrt, etc.)*
92+
- **Missing Value**: Default value if field is absent
93+
- **Range Conditions**: Define value-based thresholds (`>`, `<`, `>=`, `<=`)
94+
95+
**Example:**
96+
Boost items where a field (e.g., `views`) is greater than a defined threshold.
97+
98+
99+
### Restore Defaults
100+
101+
Each configuration section provides a **Restore Defaults** button to revert to system defaults.
102+
103+
{% image
104+
src="/images/v1.10/how-to-guides/admin-guide/restore-defaults.png"
105+
alt="Restore Defaults"
106+
caption="Restore Defaults"
107+
/%}
108+
109+
## Best Practices
110+
111+
- Use **Term Boosting** to elevate priority metadata like `Tier` or `Tags`.
112+
- Apply **Field Value Boosting** for popularity metrics (e.g., `views`, `usage`).
113+
- Avoid **excessive boosting** — keep scoring balanced.
114+
- Use **Match Types** wisely to prevent unintended exclusions (e.g., `Exact Match` vs `Partial`).
Lines changed: 114 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,114 @@
1+
---
2+
title: Search Configuration Settings for Relevance, Ranking, and Field Boosting
3+
description: Detailed guide on customizing search relevance and behavior across a platform, including global rules, term boosting, field weighting, and asset-specific configurations.
4+
slug: /how-to-guides/admin-guide/search-configuration-settings
5+
---
6+
7+
# Search Configuration Settings
8+
9+
The **Search Configuration** section in OpenMetadata allows administrators to fine-tune the relevance and behavior of search results across the platform. You can define these settings at two levels:
10+
11+
- **Global Search Level**
12+
- **Data Asset Type Level** (e.g., Containers, Dashboards, APIs)
13+
14+
## 1. Global Search Configuration
15+
16+
These settings impact **all searches across data assets**. Use this to apply platform-wide rules for performance, security, and scoring.
17+
18+
### 🔗 Access Path:
19+
`Settings → Preferences → Search`
20+
21+
### Configurable Options:
22+
23+
| Setting | Description |
24+
|-----------------------------------|-----------------------------------------------------------------------------|
25+
| Enable Roles & Policies in Search | Toggle to include role-based access policies in search filtering. |
26+
| Max Aggregate Size | Maximum size of aggregated search results (default: `10000`). |
27+
| Max Result Hits | Maximum number of search hits returned in a single query (default: `10000`). |
28+
| Max Analyzed Offset | Controls how much of a field’s content is analyzed (default: `1000`). |
29+
30+
### Term Boost (Global)
31+
32+
Define boost values to give priority to specific field terms in the search scoring mechanism. For example, Tier levels:
33+
34+
| Field | Boost |
35+
|--------------|-------|
36+
| Tier.Tier1 | 50 |
37+
| Tier.Tier2 | 30 |
38+
| Tier.Tier3 | 10 |
39+
40+
- **Use Case:** Prioritize Tier 1 assets over lower-tiered ones in search results.
41+
42+
{% image
43+
src="/images/v1.9/how-to-guides/admin-guide/global.png"
44+
alt="Global Search Configuration"
45+
caption="Global Search Configuration"
46+
/%}
47+
48+
## 2. Data Asset Level Search Configuration
49+
50+
Search ranking can be further refined per **data asset type**, such as Containers, Dashboards, or APIs. These allow **domain-specific control** over what fields or values influence the ranking of search results.
51+
52+
### Matching Fields
53+
54+
Defines which fields contribute to determining relevance in search.
55+
56+
| Property | Description |
57+
|----------------|--------------------------------------------------|
58+
| Field | The metadata field (e.g., `name.keyword`) |
59+
| Match Type | Determines type of match (e.g., `Exact Match`) |
60+
| Weight | Numeric value defining the field’s importance |
61+
| Highlight Fields | Toggle to highlight this field in search preview |
62+
63+
**Example:**
64+
`name.keyword` set as **Exact Match** with **weight 20**
65+
66+
{% image
67+
src="/images/v1.9/how-to-guides/admin-guide/data-asset-level-search.png"
68+
alt="Data Asset Level Search Configuration"
69+
caption="Data Asset Level Search Configuration"
70+
/%}
71+
72+
73+
### Term Boost
74+
75+
Boosts specific field values to rank some records higher.
76+
77+
| Field | Boost |
78+
|--------------------|---------------|
79+
| dbtTags.model_tag_one | (Custom value) |
80+
81+
82+
83+
### Field Value Boost
84+
85+
Boost based on **numerical or date-based field values**, using conditions and modifiers.
86+
87+
**Configuration Fields:**
88+
89+
- **Field**: Choose the field to evaluate
90+
- **Impact**: Boosting factor
91+
- **Modifier**: *(none, log, sqrt, etc.)*
92+
- **Missing Value**: Default value if field is absent
93+
- **Range Conditions**: Define value-based thresholds (`>`, `<`, `>=`, `<=`)
94+
95+
**Example:**
96+
Boost items where a field (e.g., `views`) is greater than a defined threshold.
97+
98+
99+
### Restore Defaults
100+
101+
Each configuration section provides a **Restore Defaults** button to revert to system defaults.
102+
103+
{% image
104+
src="/images/v1.9/how-to-guides/admin-guide/restore-defaults.png"
105+
alt="Restore Defaults"
106+
caption="Restore Defaults"
107+
/%}
108+
109+
## Best Practices
110+
111+
- Use **Term Boosting** to elevate priority metadata like `Tier` or `Tags`.
112+
- Apply **Field Value Boosting** for popularity metrics (e.g., `views`, `usage`).
113+
- Avoid **excessive boosting** — keep scoring balanced.
114+
- Use **Match Types** wisely to prevent unintended exclusions (e.g., `Exact Match` vs `Partial`).
262 KB
Loading
121 KB
Loading
247 KB
Loading
262 KB
Loading
121 KB
Loading
247 KB
Loading

0 commit comments

Comments
 (0)