Terms

Term

Description

Instance

An instance is a set of data configurations of a user, including the data source schema, index schema, and other attribute configurations. Each instance is a search service.

Document

A document is a searchable unit of structured data. A document contains one or more fields and must contain a primary key field. An OpenSearch High-performance Search Edition instance identifies each unique document by its primary key value. If a primary key value is duplicated, the existing document is overwritten.

Field

A field is a component of a document. A field consists of a field name and field content.

Plug-in

To process data during import, the system provides several built-in data processing plug-ins. You can select these plug-ins when you define the application schema or configure data sources.

Source data

The raw data, which contains one or more source fields.

Source field

The smallest unit of source data. A source field consists of a field name and a field value. For more information about the supported data types, see OpenSearch table schemas.

Index

An index is a data structure that is used to accelerate retrieval. You can create multiple indexes for an instance.

Composite index

You can configure multiple fields of the TEXT or SHORT_TEXT type in the same index to create a composite index. For example, a forum search service needs to provide search based on the title and comprehensive search based on both the title and the body. In this case, you can create a title_search index on the title field and a default composite index on the title and body fields. Then, queries on the title_search index perform title-based search, and queries on the default index perform comprehensive search based on the title and body.

Index field

Used in the query clause. You must define index fields, which are used to perform high-performance retrieval.

Attribute field

Used in the filter clause, sort clause, aggregate clause, and distinct clause to implement features such as filtering and statistics.

Default display field

Used to display results. You can use the fetch_fields API parameter to specify the fields that are returned for each query. Note that if you configure the fetch_fields parameter in your program, it overrides the default display field configuration, and the fetch_fields setting in the program takes precedence. If you do not set the fetch_fields parameter in your program, the default display fields are used.

Tokenization

Splits documents into terms. Fields of the TEXT type are split by retrieval unit, and fields of the SHORT_TEXT type are split by single character. For example, the Chinese phrase \u6d59\u6c5f\u5927\u5b66 (Zhejiang University; decode the escape sequences to obtain the actual Chinese characters) is split into 2 terms in a TEXT field: \u6d59\u6c5f (Zhejiang) and \u5927\u5b66 (University). In a SHORT_TEXT field, it is split into 4 terms: \u6d59, \u6c5f, \u5927, and \u5b66.

term

The units produced by tokenization are called terms.

Index building

After tokenization, indexes are built so that documents can be quickly located based on query requests. The search engine builds two types of lists: inverted lists and forward lists.

Inverted index

A list that consists of mappings from terms to documents. The query clause uses this method to perform queries. Example: term1->doc1,doc2,doc3; term2->doc1,doc2.

Forward index

A list that consists of mappings from documents to fields. The filter clause uses this method, which is slightly slower than the inverted index. Example: doc1->id,type,create_time.

Retrieval

The query keywords are tokenized, and the resulting terms are used to quickly locate documents by looking up the inverted lists.

Number of retrieved documents

The number of documents obtained by retrieval.

Term

Description

Data source

The source of data. Data synchronization from Alibaba Cloud RDS, MaxCompute, and PolarDB is supported.

Index rebuilding

Rebuilds indexes. After you configure or modify the application schema or data sources, you must rebuild indexes.

Term

Description

Document capacity

The total size of documents in all tables of an instance. Field names are not counted, and field content is counted as strings.

QPS

The number of query requests per second.

LCU

A logical computing unit (LCU) is a unit that measures search computing capacity. One LCU represents 1/100 of the computing capacity of one CPU core in a search cluster.

Term

Description

Sort expression

A sort expression is used to control how searched documents are sorted. Sort expressions support basic mathematical operations, mathematical functions, and built-in functions.

Rough sort expression

Performs the first round of preliminary selection on search results. Documents are scored based on the expression and sorted by their scores.

Fine sort expression

Selects the top N documents from the first-round sort results, performs a second, more detailed round of score calculation based on the fine sort expression, and sorts the documents by these scores to produce the final order.

Result summary

Text content is usually long. When search results are displayed, only the matched parts of the content can be displayed to help users quickly understand the main content of documents.