Train the similarity solution for Demand Management to find similar demands

  • Release version: Xanadu
  • Updated July 31, 2025
  • 1 minute to read

    • Train the Similar Demands solution definition included within the Predictive Intelligence for PPM capability to find related demands when creating a demand.

    Before you begin

    Ensure that the Predictive Intelligence for Contextual Search plugin (com.snc.contextual_search_ml), PPM Standard plugin (com.snc.financial_planning_pmo), and Predictive Intelligence for PPM plugin (com.snc.ppm_ml) are activated. For more information about Predictive Intelligence, see Predictive Intelligence.

    Role required: admin

    Procedure

    1. Navigate to All > Predictive Intelligence > Similarity > Solution Definitions.
    2. In the Similarity Definitions list, search for and select the Similar Demands solution definition (ml_sn_global_global_similar_demands).
    3. On the Similarity Definition form, verify the default field values for demands.
      For more information about the Similarity Definition form fields, see Create and train a similarity solution.
    4. Click Update & Retrain.
    5. Open the Similarity Definition form for the Similar Demands solution definition (ml_sn_global_global_similar_demands).
    6. In the ML Solutions related list, view the training solution progress in the Progress column.
      Note:
      Alternatively, you can click the link for the solution in the Active column. On the ML Solution form, click the Show training progress related link to check the training solution progress.

    Result

    When the solution is complete, the similar demands appear in the Similar demands section on the Demand form.

    What to do next

    • Review similarity examples: On the Similarity Definition form, in the ML Solutions related list, when Progress is 100%, in the Active column, click the link for the solution. On the ML Solution form, click the Similarity Examples related link to view the Similarity Examples list.
    • Update the similarity score threshold: On the ML Solution form, on the Solution Statistics tab, enter the required value in the Similarity Score Threshold field. Right-click the ML Solution form and click Save.
    • Test the prediction output for the records: On the ML Solution form, on the Test solution tab, enter your text in the Short description field and the maximum number of expected results in the Top N field, and then click Run test. The results above the similarity score threshold value are displayed.