What will you be doing?
When constructing semantic domain models that can be used to support IT system interaction and information exchange, many input sources are text-based. They provide human-readable information in natural language: manuals, design documents, guidelines, standards, etc. In order to generate machine-readable and machine-processable semantic models, e.g. knowledge graphs, we need to extract relevant semantics from those domain sources. This includes relevant concepts, their properties, but also their relations and potentially axioms directly or indirectly present in the text. Current approaches to Natural Language Processing are able to dissect text and generate models that incorporate e.g. a graph of grammatical concepts and relations, but these models require a lot of additional effort to make them usable as a semantically sound domain solution. Examples of such approaches are Stanford CoreNLP, HearstPatterns, Word co-occurrence, Word2Vec etc. These current tools only solve part of the problem, but e.g. do not consider document structures, do not distinguish between relevance of concepts (all concepts are equally relevant), are not able to generate semantic relations, i.e. they do not directly provide a useful domain abstraction. This assignment is targeted to further develop these methodologies in order to extract better knowledge graphs for a specific domain, especially focusing on the Smart Industry and AgriFood domains.
The purpose of this research is to improve on our engineering approach that creates a useful semantic model from textual input sources. This work will be part of various projects within TNO in which collaboration with partners in the Smart Industry and AgriFood sector, such as Philips or Unilever and FrieslandCampina, is possible and several use cases can be provided in that context.
At TNO, there is already a first version of the Semantic Ontology Engineering Toolset (SONNET) platform available containing some of the NLP tools to generate semantic models. This toolset needs to be extended with:
- Tools that generate a semantic model from natural language by taking both a-priori document knowledge (structure, type, etc.) and document characteristics (e.g. concept occurrence and co-occurrence statistics, spatial/position measures, etc.) into account.
- Tools that can improve on generated semantic models that takes heuristic rules into account.
- Extension of an existing framework for measurement of the quality of a semantic model as well as the aspects that determine this quality, such as the domain, the purpose or the application of the semantic model.
- Methods and corresponding tools that can be used to evaluate the quality of the semantic models generated by the integrated NLP tools.
We want you to contribute to this!
What do we require of you?
Your background and interests should include natural language processing, semantic modeling and artificial intelligence. We expect from you to have a creative way of thinking to come up with new solution directions and innovations that are out-of-the-box and not yet thought of in the context of the problem to be solved. In addition, you are a teamworker that is also interested in the work of others in the department and has an open eye for related work that can be applied to the problem at hand. Finally, a result-oriented attitude is necessary for this assignment and it is expected that a working prototype of the proposed solution will be realized.
What can TNO offer you?
You want to work on the precursor of your career; a work placement gives you an opportunity to take a good look at your prospective future employer. TNO goes a step further. It’s not just looking that interests us; you and your knowledge are essential to our innovation. That’s why we attach a great deal of value to your personal and professional development. You will, of course, be properly supervised during your work placement and be given the scope for you to get the best out of yourself. Naturally, we provide suitable work placement compensation.
Has this vacancy sparked your interest?
Then please feel free to apply on this vacancy! For further questions don’t hesitate to contact us.
Note that applications via email and third party applications are not taken into consideration.
Contact: Jack Verhoosel
Phone number: +31 (0)88-86 62459