Home » Project 2 – Artificial Intelligence for Data Mining

Project 2 – Artificial Intelligence for Data Mining

The second of 3 analytical consortium projects in the TIES Living Lab. This project is managed by Lamine Mahdjoubi of University of the West of England

The purpose of this project is to develop a semi-automated Artificial Intelligence (AI) driven system, that facilitates cost data mining for benchmarking. Using advanced AI methods and techniques, this project will enable the efficient extraction and translation of cost information produced in different formats and using different methods of measurement, into an agreed common format such as the International Construction Measurement Standard 2 (ICMS) or other appropriate standard structures determined by partners. enabling comparable cost metrics and benchmarks to be consistently determined and will also provide a robust basis for other metrics based on costs and quantities, such as environmental benchmarks and whole life analysis.

The project promises to:

  • Build a data pre-processing pipeline to gather and prepare cost data into an appropriate format for further analysis using artificial intelligence (including machine learning techniques).
  • Develop and train an array of Artificial Intelligence (AI) techniques to classify cost data descriptions into a common standard.
  • Carry our iterative testing of the accuracy of the classification obtained from the AI system and feed the information back into it in an ongoing learning process.
  • Develop the approach into a semi-automated system for extraction, identification, translation and mapping of cost related data into a standardised schema for further benchmarking analysis (e.g. ICMS or other appropriate formats for partners).

 

Project Outcomes

  • Tested semi-automated AI systems for mining cost data (i.e. cost data extraction, identification, translation and mapping into standardised schema).
  • Deliver cost and quantity data classified in an agreed common standard (e.g. ICMS or other standards appropriate for the emerging data) for use in Project 3 for benchmarking analysis to support decision making