Effective Master Data Management (MDM): what, why, how
summary
HAPI MDM provides a three-stage workflow: candidate filtering to limit pre-searches, applying comparisons and algorithms to score pairs, and composing a match result map to decide matches and possible matches. The module supports any resource types, interceptors, custom search params and leaves survivorship to a separate layer (Smile). Machine learning automates threshold selection, prunes and rewrites result maps via a prefix-trie decision-tree optimizer, and uses a voting ensemble of logistic regression random forest and Gaussian NB; ML-tuned configs outperform manual tuning in initial tests.
AI-written summary · about
chapters (9)
- 0:00Introduction to MDM
- 1:48What is Master Data Management
- 3:11Why MDM matters in healthcare
- 5:02Tuning MDM rules for performance
- 12:18Applying comparisons and algorithms
- 19:19Machine learning in MDM
- 27:06Why use ML for MDM
- 33:25Q&A session begins
- 47:00Closing remarks
related talks
53mBringing Mautic and Drupal closer together in an open DXP
32mAdvanced on-device ML made easy
24mSearch and find. How we made MDN discoverable
12mMachine-Generated Data and SAP's Big Data Analytics Portfolio
47mMapping FHIR to OMOP Using Open Source Tools
38mSeamless Integration of Experience Manager Assets with Applications
37mIHE Profiles and FHIR (2)
20mThe AI-Driven DXP: New Horizons for Marketers