Ground Truth Jobs

Configuration Properties

Estimate ground truth queries using click signals and query signals, with document relevance per query determined using a click/skip formula.

Tip	Pair this job with the Ranking Metrics job to calculate relevance metrics, such as nDCG.

Configuration Properties

Property Description, Type

filterQueries

Filter Queries

Filter queries to apply while choosing top queries from query signals in signals collection

type: array of string

Spark Job ID

required

The ID for this Spark job. Used in the API to reference this job

type: string

maxLength: 128

pattern: ^[A-Za-z0-9_\-]+$

joinKeySearchLogs

Join Key (Query Signals)

Join key of query signals in the signals collection

type: string

default value: 'id'

joinKeySignals

Join Key (Click Signals)

Join key of click signals in the signals collection

type: string

default value: 'fusion_query_id'

searchLogsAddOpts

Additional Options

Additional options to use while loading search logs collection

type: object

object attributes: {
}

searchLogsPipeline

Search logs pipeline

Pipeline id associated with search log entries

type: string

minLength: 1

signalsAddOpts

Additional Options

Additional options to use while loading signals collection

type: object

object attributes: {
}

signalsCollection

Signals collection

required

Collection containing click signals and the associated search log identifier

type: string

minLength: 1

topQueriesLimit

Top Queries limit

Total number of queries to pick for Ground truth calculations

type: integer

default value: '100'

type

Spark Job Type

required

type: string

default value: 'ground_truth'

enum: { ground_truth }