Internship: Part replacement identification using Knowledge Graphs
Description
(This project is also available as an MSc Thesis)
Company: Marel
Location: Boxmeer
Background
It is important for industrial equipment developers to provide accurate part replacements to their customers. Parts can wear over time or break and having suitable replacements is a dynamic process based on availability, machine version, specifications and regional differences. Knowledge Graphs (KG), which structure data by connecting entities and their relationships, offer a comprehensive view of parts and their attributes. Knowledge Graphs are a promising solution to this problem. By leveraging this structured information, the system could efficiently identify and retrieve the correct replacement parts based on specific criteria, enhancing the accuracy and speed of part identification.
Proposal
This project aims to research retrieval methods in order to retrieve a fitting and suitable replacement part from a knowledge graph, based on previous specifications. It is important to closely align the solution with the needs of the end users at technical support.
Goals
- Investigate the characteristics for similar node retrieval from KG's in an industrial setting
- Build a proof-of-concept, based on end user requirements serving in technical support.
- Verify the effectiveness and usability of the solution, both from the data mining and end-user perspective
Student requirements
- Having passed (or intending to take) Knowledge Engineering (2AMD20) is a pre
- Pragmatic but research oriented, with the intention to publish the results
- You have experience or are interested in working with natural language processing techniques
- Work in Boxmeer 3-4 days a week
Perks
- An internship allowance and travel allowance
Details
- Supervisor
-
Mykola Pechenizkiy
- Company
- Marel
- Interested?
- Get in contact
- Link
- Thesis