Oppgaven er ikke lenger tilgjengelig

Data Lake (Schneider Electric)

Schneider Electric Norway AS has proposed master projects that will be co-supervised by them and will involve real and challenging ICT problems in their organisation. Part of the work will be done at their offices at Ryen. 

Schneider Electric Logo

Schneider Electric Logo

Data lake

Description of Problem

Schneider Electric have acquired many different companies through the years and the product portfolio and the complexity in systems and information layers have grown with it. With the different layers of product information – whether it is analogue or digital, the challenge is to have a data lake with all products (legacy and commercialized), global and localized data, and to have the data structured. This would be an enabler for to drive to meet the customers expectation and the need for information leading to a conversion.  

Today we have challenges:

  • To find information on legacy product & and many instances the information is not digital available
  • manual processes as localizing/translation of technical information
  • To automate the delivery of product data
  • To dynamically connect keywords and optimize description for products

Scope of Project

  • Merge data bases and product related data to make a data lake for all product (related) data.
  • Structure data and enable easy distribution of the data
  • Gather user data from own web and search engine data.
  • Merge data lake and user data to create/ generate keywords, SEO product description, product relations,

Practical Details

Examples of systems to find product data: ERP, PIM, PDF, product selectors, handwritten, excel, BOX, printed catalogues. 3rd party data bases (efobasen.efo.no, bimobject.com, traceparts.com)

This project is in the areas of databases, semantic technologies and data science. The project’s theoretical contribution will be to determine an effective method of governing and describing the product information in the Schneider organization. Application of ontology-based data access methods to this problem should be addressed. The project fits naturally into the area of interest of the Knowledge Representation group in SIRIUS.

Emneord: data science, Big Data, semantic technologies, Statistics
Publisert 5. okt. 2017 09:12 - Sist endret 16. mai 2018 09:47

Omfang (studiepoeng)