Beschreibung
Aufgabe:We are looking for someone with expert level knowledge in text mining & natural language processing to support a project focussed on information retrieval, information extraction and knowledge discovery from unstructured data sources such as patents and social media (data sources could be extended). He/She is expected to play a dual role 40:60 of business analyst & text mining specialist/developer. The project will be bundle of pilots for specific business sectors within our organization. Success criteria of these pilots will define "operationalize" or "stop" for individual pilots.
- Analysis of business requirements
- Prepare scope and high level solution design (functional & technical)
- Manage and educate end users for mainly unrealistic expectations
- Able to professionally communicate during individual scoping meetings or team meetings incl. presentations
Hands-On:
- Able to build quick pilot solutions (ranging 2-4 weeks) using existing inhouse applications or using open sources
- Distribute task/work to offshore resource(s) [data science profile] who could support you with hands-on build, optimize, test activities
Anforderung:
- Expert level knowledge in natural language processing, information retrieval, information extraction & statistical modeling
- Hands-On experience in named entity recognition, text classification (Documents or Sentences), key phrase identification, sentiment analysis (incl. ability to train ML models if required)
- Dealt with thesauri, controlled vocabularies, ontologies in formats such as XML, RDF which will be used a backbone for named entity recognition
- Able to parse source data which will be in XML or JSON formats.
- Expert level knowledge in Python packages for NLP & Machine Learning (ML) (i.e. NLTK, Spacy, ScikitLearn, etc)
- Experience with document indexing engines such as Lucene, Solr or Elasticsearch
- Experience working in big data environment and distributed computing i.e. Hadoop & Spark is desired
- Knowledge of deep learning applied to text data is welcome BUT not mandatory
- Knowledge of analysis tools such as Kibana or QlikSense is welcome
Umgebung/Sonstiges:
- Utilization per Week (Days): 4
- On-site Support (in Percentage): 100%
Beginn: Juli 2018
Dauer: 14.12.2018+
Branche: Chemie/Pharma