Persnl Rcmd Item Log DMO

Stores per-item decision signals for each candidate evaluated or returned within a recommendation inference event. As the child of Recommendation Log, each record captures propensity scores, posterior distribution parameters, exploration flags, and item role for a single item in the ranked response. Downstream components like the Contextual Bandit Analytics Engine consume these records to compute per-item performance metrics - cumulative regret, IPS-weighted calibration, and off-policy evaluation.

Object API Name: std__PersnlRcmdItemLogDmo__dlm
Category: Unassigned
Availability: Available in 262 and later versions
Primary Key Field: Id

• PersnlRcmdLogId has a FOREIGNKEY relationship with the Persnl Rcmd Log DMO Id field.
• IndividualId has a FOREIGNKEY relationship with the Individual DMO Id field.
• PersonalizerId has a FOREIGNKEY relationship with the Personalizer DMO Id field.

• cdp_sys_record_currency
• DataSourceId
• DataSourceObjectId
• Id
• IndividualId
• InferencePolicyType
• InternalOrganizationId
• IsForcedExplore
• ItemIdentifier
• ItemRoleType
• ItemType
• PersnlRcmdLogId
• PersonalizationIdentifier
• PersonalizerId
• PropensityScoreNbr
• RelevanceMeanNbr
• RelevanceScoreNbr
• RelevanceVarianceNbr
• RequestDateTime

• Field API Name: std__cdp_sys_record_currency__c
• Data Type: TEXT
• Description: System-generated metadata field that stores a 3-letter ISO currency code.

• Field API Name: std__DataSourceId__c
• Data Type: TEXT
• Description: A unique, system-generated identifier for this record.

• Field API Name: std__DataSourceObjectId__c
• Data Type: TEXT
• Description: A unique, system-generated identifier for this record.

• Field API Name: std__Id__c
• Data Type: TEXT
• Description: System-generated unique identifier for this per-item decision record within a recommendation inference event. Serves as the row-level primary key — distinct from the parent Personalization Recommendation Log Id, which identifies the inference request that produced this item.

• Field API Name: std__IndividualId__c
• Data Type: TEXT
• Description: Reference to the Individual in whose context this recommendation was produced.

• Field API Name: std__InferencePolicyType__c
• Data Type: TEXT
• Description: Indicates the policy phase of the contextual bandit model (Machine Learning model) at decision time - for example, whether the model was in cold-start phase or fully-trained phase.

• Field API Name: std__InternalOrganizationId__c
• Data Type: TEXT
• Description: A unique, system-generated identifier for this record.

• Field API Name: std__IsForcedExplore__c
• Data Type: BOOLEAN
• Description: Indicates whether this item was selected via forced random exploration (epsilon-Thompson hybrid) rather than model-driven sampling.

• Field API Name: std__ItemIdentifier__c
• Data Type: TEXT
• Description: Identifier of the specific item this row represents within the recommendation response. Uniquely identifies the returned item (e.g., a product, an offer, a channel) within the contextual bandit model’s (Machine Learning model) action space.

• Field API Name: std__ItemRoleType__c
• Data Type: TEXT
• Description: Indicates the role of this item within the inference decision: whether it was the recommended item that was chosen and served to the caller, or a candidate item that was eligible but not selected.

• Field API Name: std__ItemType__c
• Data Type: TEXT
• Description: The classification or kind of item this row represents, depending on the recommendation use case.

• Field API Name: std__PersnlRcmdLogId__c
• Data Type: TEXT
• Description: Identifier of the parent Recommendation Log record - the originating inference request that produced this returned item.

• Field API Name: std__PersonalizationIdentifier__c
• Data Type: TEXT
• Description: A system-generated UUID assigned to each personalization decision produced by the personalization pipeline. A single inbound request may target multiple personalization points on a page (for example, five different surfaces personalized in one call), and each point receives its own Personalization Id.

• Field API Name: std__PersonalizerId__c
• Data Type: TEXT
• Description: Identifier of the recommender (model configuration) that produced this inference, scoped to the tenant. Distinguishes inference events across different recommenders within the same tenant and allows training pipelines to narrow the dataset to the recommender being retrained.

• Field API Name: std__PropensityScoreNbr__c
• Data Type: DOUBLE
• Description: The probability that the contextual bandit model (Machine Learning model) policy would have selected this item given the context and candidate set at decision time.

• Field API Name: std__RelevanceMeanNbr__c
• Data Type: DOUBLE
• Description: Mean of the posterior distribution assigned to this item by the contextual bandit model (Machine Learning model) at decision time.

• Field API Name: std__RelevanceScoreNbr__c
• Data Type: DOUBLE
• Description: Stores the contextual bandit model (Machine Learning model) assigned score reflecting how relevant this candidate item is to the individual’s context at decision time.

• Field API Name: std__RelevanceVarianceNbr__c
• Data Type: DOUBLE
• Description: Variance of the posterior distribution assigned to this item by the contextual bandit model (Machine Learning model) at decision time.

• Field API Name: std__RequestDateTime__c
• Data Type: DATETIME
• Description: Timestamp at which the recommendation engine began processing this inference request.