New Year, New Tool. Introducing Our New Classifier.

January 31, 2021
Procurement

Our New Solution To Classification

In 2020, we developed an advanced classifier. This tool adds multiple labels to procurement notices based on the Common Procurement Vocabulary (CPV), with an accuracy of 84%. This classifier gives notices five scored, Level 3 CPV codes based on their text and description, that work for the 90 languages in our database. This classifier can process millions of recommendations within hours. As of writing, we are classifying all our new records each day, with full predictions appearing in our API in early in 2021.

This classifier is a series of algorithms that were trained on millions of rows of procurement data in our database. Our model has an F1 score of at least 84%, a precision of 83% and a recall of 88%.

What Does F1, Precision and Frequency Mean?

These terminologies are industry standard terms within the practise of demonstrating algorithmic accuracy.  F1 score is the ‘harmonic’ mean or average of precision and recall. Precision analyses correct and incorrect predictions based on the rows that went into the model. Recall looks at the wider picture: which rows the model got right versus the ones it completely missed.

The Challenges With The Classification

Developing code to categorise spending and tenders has proven technically challenging, due to a number of complicating factors in the underlying published data and the scale of the problem to be addressed.

Multi-Label Data Challenge. All of the documents published in TED and many of the public sources around Europe publish documents with multiple classifications, with no primary classification identified. Many records have more than ten classification references and some buyers deploy a technique of adding diverse requirements to a single tender in different lots. This requires buyers to provide diverse categorisations for the same tender, making it impossible to identify a primary classification for a tender.

So a document with the following codes 72541000, 72510000, 72590000, 30210000 would be classified to 72000000 : IT Services: consulting, software development, Internet and support.

Inaccurate Classification Challenge. We’ve found that a high number of documents have been incorrectly classified. Mostly this is down to confusion around the correct class to apply, but occasionally there is a clear misapplication of codes, usually around services. We often see consulting described as training or research. This problem is exacerbated by the inconsistent nature of CPV codes which are poorly optimised for the description of services. ‘Consultancy’ is mentioned 6 times in the top two layers of the classification and in four different divisions of the vocabulary.

Poor Narrative Data. With some documents having just five words in their combined titles and descriptions, this is barely enough to classify the documents and makes the task of classifying data much harder. This issue is particularly prevalent in Spain where narratives are, on average, half as long as narratives in the UK.

Scale of Classification Challenge. Scale increases the complexity of the analysis we need to do significantly, with thousands of classification options open to the algorithm. The risk of inaccurate classification increases exponentially with the number of classes to be used.

Addressing These Challenges

With our classifier tool, we tackle each of these problems. Our tool is detailed, accurate, and can be deployed at scale. We can use this on private and public sector tenders, contracts and spend. This tool classifies to CPV but we can adapt this tool to other major classifications.

Using The Classification

→ In our api / source data

We can provide data with enhanced classification through our off-the-shelf API or through a custom API. This can be done for either all or a curated part of our data. For instance, if you are a haulage company, we can give you a customised feed of tender, contracts and spend data for logistics, infrastructure, construction.

→ We can classify your data, send and return

We can take your dataset, apply classification labels and send it back to you. Whether you want more details of corporate costs or a B2C analytics firm looking at better segmentation of consumer spend, we can help.

→ Building a custom classification for you

We can work with existing classifications such as CPV or Proclass or even make one for you, tailored to your needs. We can create a new classification that both fully reflects modern buying trends such as consultancy, and improves on existing, outdated taxonomies such as digital and software services.

Examples

Where We Improve The CPV Level

TITLE ORIGINAL CPV PREDICTED CPV
GP IT Systems IT services: consulting, software development, Internet and … Software-related services
Provision of Summative Assessment … Business and management consultancy and related … Business and management consultancy services
Re-Roofing of Admin Building, Police HQ Portishead BS20 8QJ Construction work Roof works and other special trade construction works
MPV Passenger Transport without Passenger Assistance Transport services (excl. Waste transport) Taxi services

 

Adding CPVs Where None Existed

TITLE PREDICTED CPV
New build mixed-use development of 17 apartments and business hub Building construction work
PHE – ICT – Application Service Support and Maintenance for contract tracing solution Software-related services
Citizen & Consumer Protection Officer x4 Supply services of personnel inc. temporary staff
Provision of Youth and Young Carers Services Social work services
Evaluation of the Small Business Leadership Programme (SBLP) Business and management consultancy services

Different CPV For Better Accuracy

TITLE ORIGINAL CPV PREDICTED CPV
The Provision and Maintenance of NHS Health Check Software Health and social work services 85000000 Software-related services 72260000
C-19 Corporate Finance and Market Intelligence Support… Business and management consultancy … 79400000 Financial consultancy… 66170000
CA7688 – Roofing Services Tender Building services 71315000 Roof works and other works 45260000
Housing Solutions Officer x11 Housing services 70333000 Supply services of personnel inc temporary staff 79620000

Multilingual Example

TITLE ORIGINAL CPV PREDICTED CPV
Puhastusteenused Cleaning services Cleaning services
Services d’assurance Insurance services Insurance services
Villamos energia beszerzése Electricity Electricity
Furnizare alimente Chickens Meat
Rahmenvereinbarung Lieferung und Montage von Schulmöbeln School furniture School furniture
Servicios de reparación y mantenimiento … Repairs & Maintenance  Repairs and Maintenance

if you’d like:

  • classified procurement data;
  • to have your own data classified;
  • or to build your own classifier.
January 24, 2023

Improve Your Exporting With Government Data

Government procurement data can be a valuable resource for export agencies looking to expand their business and identify new opportunities in foreign...
January 18, 2023

Waiting For A Tsunami Of Text

In a world where text can be generated by an algorithm for negligible costs, we have to rethink how we value the...
January 17, 2023

Cleaning Up On Cleaning Contracts

Canberra, in the Australian Capital Territory, is home to the nations houses of parliament and a large swathe of public sector. Transparency...
January 11, 2023

New Procurement Bill Progresses Through Parliament

The proposed new Procurement Bill passed its second reading without division in the House of Commons yesterday, with the Government proposing the...
January 10, 2023

How Government Procurement Data Can Help Export Agencies

Government procurement data, the information on what goods and services government entities purchase in different markets, is an extremely valuable, and often...
January 5, 2023

Government On Covid 19 Contracts

The government has published a document, providing further information about the procurement of critical testing equipment and services during the early months...
January 3, 2023

Global Analysis: Less Is Less

Analysing 7.5 million tender documents published by governments around the world has revealed a worrying trend of publishers providing less data for...
December 1, 2022

Using Data To Build A Supply Chain

All data has a provenance. It comes from somewhere. Someone created it. Maybe it came from a sensor, or it was created...
December 1, 2022

Get Your Data Sorted.

We collect tens of thousands of documents every day. We visit more than 700 sources to collect this data. That's what our...
December 1, 2022

The Importance of Government Procurement Data for Export Agencies

There are a number of ways that government procurement data can help export agencies, including: Identifying New Markets - By understanding what...
December 1, 2022

Lord’s Have Their Say On Procurement Bill

The proposed changes to the Procurement Bill are on their third reading in the House of Lords. Last week the House of...
December 14, 2022

Buyers who stole Christmas (again)

It’s time to shed a festive tear and feel some sympathy for all the bidding teams who’ve got bids to respond to...
December 1, 2022

How Government Procurement Data Can Help Export Agencies.

If you’re in the business of exporting, you know that the global market can be a tough one to crack. In order...
December 1, 2022

EU’s Ruling On Beneficial Ownership

Last week, the Court of Justice of the European Union (CJEU) ruled that public access to registries showing personal details about companies'...

Newsletter

Compelling research, insights and data directly into your inbox.

Recent media stories

Search