NLP Expert Offline

We, a small Dutch think tank working in the area of international relations and security, have collated a number of multilingual academic full-text corpora. They were downloaded (with metadata) from various academic databases like ProQuest, EBSCO, Google Scholar, Microsoft Academic, etc. We want to apply a number of both unsupervised (especially topic modelling) AND semi-supervised NLP tools (preferably based on active learning) on these corpora to (ultimately) build rich ontologies and knowledge graphs. So we’re talking about various semantic extraction (regular ’named entities’, but also temporality-, causality-, hypothesis-, event-, sentiment-, etc. extraction) AND linking tools — all of them preferably based on more recent context- (and not just BoW-)based language models like BERT, ELMo, ULMFiT, GPT-2, etc. Any of the thus extracted ’tuples’ would then be stored in some graph-database format that can be queried in intuitive ways.

We have a team of junior and senior subject-matter experts who are able and willing to start annotating these text corpora. But what we are looking for is an experienced NLP-specialist who can advise us on which tools/algorithms to use, how to get our data in these tools, how to build models, how to validate them, how to store the findings from the models in knowledge graphs, how to visualize the results, etc.