PhD position in Large Language Models for Knowledge Harvesting, Sharing, and Management

Knowledge is a vital asset for individuals, organizations, institutions, and society. Large Language Models (LLMs) have the potential to revolutionize how knowledge is generated, shared and managed by automating the extraction of relevant information from vast amounts of unstructured data, synthesizing it into coherent narratives, and injecting it in multimodal user interactions. Recent advances, such as Retrieval Augmented Generation (RAG), enable LLMs to parse through extensive documents, identify key points, and summarize content, rendering the information more accessible and comprehensible for users. Through dialectic techniques, LLMs can facilitate (tacit) knowledge extraction from expert users and instrument knowledge sharing with novice ones. LLMs can also be fine-tuned to categorize, tag, and retrieve information, responding to queries with relevant documents or even specific text passages. Using an intermediate layer, LLMs can translate natural language querries into a compatible format for popular databases (e.g., MySQL, Neo4j) thus expanding existing knowledge bases. By enabling more efficient search and discovery processes, LLMs help maintain an organized and dynamic knowledge base, supporting continuous learning and decision-making within organizations and beyond. This PhD research will also investigate the technical and algorithmic challenges involved in information retrieval techniques beyond RAG. This research will seek to answer the following Research Questions (RQs):    
 

• What methodologies can be developed to ensure LLMs effectively summarize complex knowledge without losing critical information?
• How can LLM-generated natural language explanations enhance knowledge sharing among individuals with varying levels of expertise?
• How can continuous learning be integrated into LLM-powered systems to adapt to evolving data and knowledge without extensive retraining?
• What are the potential pitfalls when utilizing LLMs to manage knowledge (e.g., biases, hallucinations), and how can they be mitigated to ensure fair and unbiased information dissemination?

 
The successful candidate will design, develop, and evaluate LLM-powered prototypes with the aim to discover, harvest, share and manage knowledge in a context-agnostic fashion. Novel LLM-powered prototypes will be deployed in real-life contexts that encompass creativity and design, healthcare, manufacturing, energy and others to showcase (societal) impact. The successful candidate will be based at TU Delft and will have the chance to work with partners across TU Delft (e.g., Faculty of EEMCS & 3ME), beyond TU Delft (e.g., Erasmus MC, Leiden University), as well as with international and industrial project partners (e.g., Ortec B.V.).  

Requirements:

Basic Requirements (must-have):

• An MSc degree in a science field relevant to the PhD research described above (Conversational AI, LLMs, HCI, AI/ML, Computer Science).
• Demonstrated competencies in one or more of these domains: AI, HCI, LLMs, Computer/Data science, or another relevant field.
• Excellent programming, scripting, and software prototyping skills (e.g., Java, Python, C++, PHP, MySQL, Neo4j) as well as interest in further developing your design, problem-solving, programming, and analytical skills.
• Strong interest and experience in conducting empirical research (e.g., user studies), knowledge in inferential statistics and qualitative methods (e.g., content analysis) and understanding of user-centred design methods.
• Ability to work in a team, elicit project requirements, take initiative, obtain results, be systematic and communicative.
• Proficient in verbal and written English communication.

 
Additional Requirements (highly desirable):
•  Interest and experience in teaching and guiding students.
Doing a PhD at TU Delft requires English proficiency at a certain level to ensure that the candidate is able to communicate and interact well, participate in English-taught Doctoral Education courses, and write scientific articles and a final thesis. For more details please check the Graduate Schools Admission Requirements.

Salary Benefits:

Doctoral candidates will be offered a 4-year period of employment in principle, but in the form of 2 employment contracts. An initial 1,5 year contract with an official go/no go progress assessment within 15 months. Followed by an additional contract for the remaining years assuming everything goes well and performance requirements are met.
Salary and benefits are in accordance with the Collective Labour Agreement for Dutch Universities, increasing from € 2770 per month in the first year to € 3539 in the fourth year. As a PhD candidate you will be enrolled in the TU Delft Graduate School. The TU Delft Graduate School provides an inspiring research environment with an excellent team of supervisors, academic staff and a mentor. The Doctoral Education Programme is aimed at developing your transferable, discipline-related and research skills.
The TU Delft offers a customisable compensation package, discounts on health insurance, and a monthly work costs contribution. Flexible work schedules can be arranged.
For international applicants, TU Delft has the Coming to Delft Service. This service provides information for new international employees to help you prepare the relocation and to settle in the Netherlands. The Coming to Delft Service offers a Dual Career Programme for partners and they organise events to expand your (social) network.

38 hours per week

Mekelweg 2