Textkernel Talks - Neo4j usage in Textkernel
1 like610 views
Textkernel, founded in 2001, focuses on natural language processing and machine learning for the HR industry, employing 65 people and collaborating with universities. The company offers a web taxonomy editor utilizing Neo4j, with significant production statistics and benefits for improving data quality and matching services. Additionally, Textkernel hosts talks on advanced topics in technology, including Elasticsearch and graph databases.
Textkernel Talks - Neo4j usage in Textkernel
- 2. Introduction
- 3. About Textkernel • Founded in 2001, currently 65 employees • R&D spin-off • Natural Language Processing and Machine Learning research groups • Universities of Tilburg, Antwerp and Amsterdam • Specialised in information extraction, web mining and semantic searching & matching in the HR industry
- 4. About Textkernel • Founded in 2001, currently 65 employees • R&D spin-off • Natural Language Processing and Machine Learning research groups • Universities of Tilburg, Antwerp and Amsterdam • Specialised in information extraction, web mining and semantic searching & matching in the HR industry • We’re hiring! → http://www.textkernel.com/jobs/
- 5. Textkernel Talks • September: Resiliency in Elasticsearch • October: Deep Learning, the recent years • December: Introduction to Graph databases • Ideas? → talks@textkernel.nl
- 6. Web Taxonomy Editor An example of Neo4j usage in Textkernel
- 7. Your presenter: Alexey Shevchenko • Joined Textkernel in 2015 as PHP developer • Member of Jobfeed team • Participates in Web Taxonomy Editor circle • Has 5+ years in production development • Docker enthusiast
- 8. Web Taxonomy Editor • Web service that is designed to edit taxonomies • Taxonomies are stored in Neo4j as an individual trees • Used by customer to improve their companies, professions and skills taxonomies • Planned to be used by Textkernel on our own data
- 9. Production stats • 1,114,942 nodes • 1,667,364 relationships • 1.6 Gb storage • 2.4 Gb RAM
- 11. TK graph usage - professions
- 12. Relations between similar taxonomies
- 13. Benefits for using Neo4j for TK Data • Significantly improve quality of Match! service • Significantly reduce amount of work for Data & Quality team
- 14. Using Docker in Development
- 15. Questions?