Product News

Enrich

Part 2: Expense categorisation across Basiq products

Basiq’s categorisation engine uses multiple sources to categorise and enrich transaction banking data with multiple attributes. To provide industry standard categorisation of expense transactions we use Australia Bureau of Statistics (ABS).


Categorisation of expense transactions

Basiq’s categorisation engine uses multiple sources to categorise and enrich transaction banking data with multiple attributes.  

To provide industry standard categorisation of expense transactions we use Australia Bureau of Statistics (ABS).

Zero human involvement

The powerful categorisation engine takes bank transactions which are typically noisy and difficult to categorise and presents a result providing your customer with context and accuracy.  

  • no humans are used during enrichment
  • 100% machine learning classification
  • rich training data updated daily
  • multiple and varied reliable data sources
  • multi-phased classification process
  • each categorisation is scored and the machine keeps learning to improve results

Over 52 different data sources in total

  1. ABR (Australian Business Register) – Current & historical businesses including individuals, sole traders and companies
  2. Self sourced – major stores, chains, franchises, schools, physicians, service stations and more
  3. Third party sources – directory listings, menu-log, yelp and more

Classification systems

There is the option to use the any of following industry standard classification systems for expenses: ANZSIC, MCC, HEC, CPICC and COICOP

  • Australian and New Zealand Standard Industrial Classification (ANZSIC)
  • Merchant Classification Code (MCC)
  • Household Expenditure Classification (HEC)
  • CPI Commodity Classification (CPICC)
  • Classification of Individual Consumption by Purpose (COICOP)

How do I get it?

Connect

Our flagship product provides the ability to acquire financial data from +60 financial institutions – both banks and superannuation funds.

  • Each transaction is categorised using ANZSIC expense classification at the group level

Example:

DescriptionExample
TransactionKOOKAI PTY LIMIT WARRINGA NS AUS Card xx0002 27.50
ANZSIC group code425
ANZSIC group descriptionClothing, Footwear and Personal Accessory Retailing

Enhance

Use this service to further enrich and categorise bank transaction records. It allows you to derive valuable insights from your customers financial data encapsulated as three objects: category, entity and location.

  • Category provides ANZSIC expense categorisation returning the full 4 level hierarchy per transaction.

Example:

DescriptionExample
TransactionKOOKAI PTY LIMIT WARRINGA NS AUS Card xx0002 27.50
Friendly merchant nameKOOKAI
ANZSIC division codeG
ANZSIC division descriptionRetail Trade
ANZSIC subdivision code42
ANZSIC subdivision descriptionOther Store-Based Retailing
ANZSIC group code425
ANZSIC group descriptionClothing, Footwear and Personal Accessory Retailing
ANZSIC class code4251
ANZSIC class descriptionClothing Retailing

Discover

Use our Affordability report to get a full view of your customers financial health including assets, liabilities, income and expenses.

The Expenses Summary provides returns a summary of expenses by category by month.  Transaction data  is enriched and classified using Household Expense Classification (HEC) from the Australian Bureau of Statistics (ABS)

  • Most of the approximately 700 items included in the classification relate to expenditure on goods and Services
  • HEC comes with four levels (Division > Subdivisions > Groups > Classes) so Basiq provides the report at the Group level allowing you to roll-up and drill-down as required
  • use this for credit decisioning to identify spend on utility bills, rent, groceries and discretionary spend such as takeaways, gambling and clothing

Read more about Basiq’s machine learning classification engine in Part 1: What is Expense categorisation? https://basiq.io/basiq-categorisation-part-1/