Analysis request

This is the reference for the top-level properties of the analysis request JSON.

For a step-by step tutorial in building Lingo4G analysis requests, see analysis JSON overview.

Lingo4G analysis request is a JSON object with the following properties. All properties are optional.

{
  "name": null,
  "comment": null,
  "variables": {},
  "components": {},
  "stages": {},
  "output": {
    "stages": [],
    "request": false,
    "progress": true
  },
  "useCache": true
}
  

name
comment

Human-readable name and description of the request. Useful if you maintain a request repository in your project.

variables

Definitions of primitive-valued variables you can reference from other parts of the request.

Default value: {}.

components

Components: reusable configuration fragments, such as search queries or field name lists, you can reference from other parts of the request.

See Stages and components overview for the list of all available analysis components.

Default value: {}.

stages

Stages: the operations to perform as part of the request. Each entry in the map defines one operation, called an analysis stage. Every stage produces a result, such as a list of documents or labels, that can can serve as input for another stage or as the output of the request.

See Stages and components overview for the list of all available analysis stages.

Default value: {}.

output

stages

An array of stages whose results to include in the output. If not provided, Lingo4G outputs the results of all stages.

Some stages, such as matrix:​*, produce large outputs that are not required to interpret or visualize the analysis result. In such cases, use the stages array to output the results of the desired stages only.

Default value: null.

request

Output the resolved request JSON. The resolved request JSON includes:

• All properties of all stages and components.

• All default components.

• Auto-references resolved into explicit references.

• Variable references substituted with values.

progress

Indicates whether to output stage processing times and performance statistics.

Default value: false.

use​Cache

If true (default), before computing the result of each request stage, Lingo4G checks the partial results cache and uses cached result, if available. This significantly improves the latency of processing sequences of similar analysis requests (e.g. request differing by a single property of a single stage).

If false, Lingo4G bypasses the partial results cache when processing this request. To benchmark a specific request or gather complete performance statistics, set use​Cache to false to ensure that Lingo4G re-computes the results of all stages.

Default value: true.