Back to catalog
Product Data Mastery

Free-Text Description Parsing & Normalization

Deliver a rule library and a CSV Golden Record extract that maps parsed MRO attributes (size, material, unit, brand, MPN) to ERP/EAM fields with field-level confidence scores.

How this course differsFor marketing and e-commerce: parsing attributes such as size, material and brand into ERP fields.

Designed by Isak Marais, Founder & CEO, AICA Data

2.5 hours
12 lessons
intermediate

Free to join. Start straight away.

What you will be able to do
  • Extract structured attributes (size, material, unit, brand, MPN) from free-text MRO descriptions using tokenization, regex and rule-based patterns and export them as fielded JSON.
  • Normalize units of measure to ISO-compatible representations, convert between imperial and metric, and attach conversion precision and uncertainty metadata.
  • Map extracted attributes to an ERP/EAM Golden Record schema and produce a CSV ready for ingestion with field-level confidence scores.
  • Evaluate parsing performance across bearings, valves and electrical components using precision, recall and error-class analysis to identify systematic failure modes.
  • Design governance rules for canonical brand lists, MPN reconciliation, and confidence thresholds to operationalize parsing at scale.

Part of the Classification Authority path

What this course proves

Completing every lesson and passing each module knowledge check at 75% first-attempt accuracy issues a verifiable AICA credential with a unique ID.

Module 1
Draft rule library (tokenization and regex rules) and a short governance brief aligning units and canonicalisation to standards.
Module 2
CSV Golden Record extract mapped to ERP/EAM fields with field-level confidence scores, plus a capstone report and deployment checklist.

Lessons in this course include hands-on tasks in the AICA Data Platform. A separate platform login is required.

Open platform

Your credential

Proof designed to travel with you

A verifiable certificate, LinkedIn share card and square badge, issued when you complete the course and pass its assessments.

Sample preview

Sample β€” yours will carry your name and unique credential ID.

Achievement levels

Pass

75–84%

Distinction

85–94%

High Distinction

95–100%

Result levels use 70% first-attempt mastery + 30% corrected completion. You must also meet the course’s 75% first-attempt pass requirement; a projected result is not an award.

Explore all credential levels

Who designed this course

IM

Isak Marais

Founder & CEO, AICA Data

Isak Marais is the founder and CEO of AICA Data, and a recognised authority on product, material and spares data. He advises boards and C-suite leaders on AI strategy, enterprise data platforms and digital transformation, and has led AI-driven product data cleansing, enrichment and governance programmes in operational use across 55 countries. He designed and authored the AICA Academy curriculum.

Product data cleansing
Data enrichment
Data governance
UNSPSC classification
Enterprise data platforms
AI strategy
  • CEO of the Year Awards 2024 β€” CEO Monthly
  • Founder & CEO, AICA Data
View LinkedIn profile

Curriculum designed and authored by Isak Marais, developed with AI assistance and reviewed against Quality Matters and OLC OSCQR standards.

Learner reviews

No published reviews yet.

Course Curriculum
1

Parsing & Normalization Foundations: Patterns, Units, Governance

6 lessons β€’ Foundations

1 hour 15 minutes

2

Applied Parsing & Operationalization: Lab, Edge Cases, Capstone

5 lessons β€’ Implementation

1 hour 15 minutes

Summative Assessment

Free-Text Description Parsing & Normalization Capstone Project

Final Assessmentβ€’60 min

Free to join. Start straight away.

Duration2.5 hours
Lessons12
Levelintermediate
Shape the Academy

Help us improve this course

Report an issue or suggest an improvement β€” every idea is reviewed.

Shape the Academy

Spotted something? Have an idea?

Your feedback shapes every course. Report an issue, suggest an improvement, or request a new topic β€” the team reviews every submission.

Β© AICA β€” licensed for enrolled learner use only; redistribution prohibited.