Pega Prediction Topic Model - Learning capability when used inside Data Flow

General
, , ,
1

As of our requirement, created a Prediction Model and used that model (Text Analyzer ) inside Data flow As below,

Data Flow with Text Analyzer .png
Data Flow with Text Analyzer .png757×244 14.7 KB

Prediction Model Description,

  • Topic Analysis : NLPTestTopic- using Maximum Entropy Model
  • Sentiment analysis - default
  • Entity Extraction : using Key Words Extraction

Prediction Model Not associated with any Channel,

Prediction Model Description .png
Prediction Model Description .png1127×610 23.5 KB

Once we Executed the Data Flow, since we are using ML(Machine Learning) for topic detection

  • Q 1. will the model self-learn, although we are using it in Data flow?
  • Q 2. Is there any additional configuration that needs to be done when we are using the text analysis model directly inside the data flow - to enable self-learning?
  • Q 3. possibility to provide feedback to the model and correct it since it is referred inside the data flow?
2

Hi @AnthonyG5648

Q1 : Pega NLP models do not self learn just by executing them. You need to provide feedback whenever it detects wrongly using feedback API and update the model.

Q2: We do not support self learning as a setting, recommendation is provide feedback and update the model.

Q3: To provide feedback : Pegasystems Documentation