Big Data

Generating structured data with informative value out of unstructured data from varying sources is a key task in the field of big data. We hold business and IT in equal regard, meaning we can optimally support our customers in these core fields of activity: filtering out junk data, effectively determining the information value of data materials, and the subsequent interpretation of the resulting meta data.

Use Case Big Data

+
Duration: 4 months Industry: Energy Function: Business Analysis

Our customer approached us with the following problem: There was a large amount of trade data being accessed by a wide variety of users and dealers. The main focus of the project was on the following points:

  1. Dealer behaviour/fraud detection
  2. Data waste

The company wanted to collect useful meta data/big data for analysis and clustering. The statistical analysis in regard to the aforementioned focus areas showed trade patterns (clusters). In addition, a list was created of commercial transactions for which there was an estimated probability of 90% that fraud had taken place.  Based on this list, the company is currently in the process of checking the list elements and analysing whether cases of fraud actually took place. Every list element that leads to a wrong prediction is regarded as a false positive by the big data solution. This information is deleted. Every element that appears in the case of actual fraud remains on the list. The remaining elements are defined as a “golden standard”. The second point of the analysis showed that only 40% of all trade data was actually used. All other data could therefore be considered junk data.

 

Amir Feghhi
Director

Amir Feghhi has more than 12 years of experience in project management, CRM, and digital and business development. He has previously worked with start-ups, medium-sized businesses and corporations, predominantly in the fields of finance, automotive, energy, the food industry and technology.