Début du message réexpédié : De: Renaud Delbru <renaud.delbru@siren.io> Objet: Offre These Cifre Date: 9 avril 2025 à 11:47:52 UTC+2 À: nicolas.boutry@epita.fr Renvoyé-De: <nicolas.boutry@epita.fr> Bonjour, Je vous joins ci-dessous l’offre. Je vous serais reconnaissant de bien vouloir la partager, aussi bien avec les anciens étudiants qu’avec ceux en fin de cycle. Merci d'avance Bien cordialement --- The Images and Contents team [1] of the L3i lab [2] at La Rochelle University invites applications for a fully-funded PhD POSITION. This PhD is funded through a CIFRE agreement in collaboration with Siren Analytics [3], allowing the candidate to work jointly between the research lab in La Rochelle and the partner company in Bordeaux. Application deadline: 21/04/2025 Knowledge Graph Construction and User-Centric Interaction for Data-Driven Investigation In the era of big data, extracting meaningful insights from vast, heterogeneous datasets remains a major challenge. Knowledge graphs (KGs) have emerged as powerful tools for structuring, linking, and visualizing complex relationships within data, enabling more effective data investigation and knowledge discovery. This thesis focuses on the construction and visualization of knowledge graphs to support data-driven investigations across various domains. The research will explore automated methods for knowledge extraction, entity linking, and relationship inference from structured and unstructured data sources. Advanced Natural Language Processing (NLP) and Machine Learning techniques will be employed to enhance the accuracy and completeness of generated graphs and to develop advanced question-answering capabilities that translate complex natural language queries into structured graph-based queries. A key aspect of the thesis will be the development of interactive visualization techniques that facilitate intuitive exploration of knowledge graphs, allowing users to navigate, query, and analyze connections efficiently. The project will address challenges related to scalability, uncertainty representation, user interpretability, and traceability, ensuring that the generated knowledge graphs are both accurate and actionable for investigative purposes. What we offer: A yearly scholarship of €40,000 is provided for the entire duration of enrollment, up to a maximum of 36 months. Collaborative environment with industry experts from Siren Analytics, offering direct exposure to real-world investigative intelligence challenges. Eligibility: Master’s degree in Computer Science, Computational Linguistics, Information Retrieval, Data Science or a related field. Strong interest or prior experience in NLP, knowledge graphs, semantic technologies, or interactive data visualization. Mode of study: Full time Year of entry: 2025 Duration: 3 years Additional criteria: Applicants must not already (i) hold a doctoral degree; or (ii) be matriculated for a doctoral degree at La Rochelle University, or another institution. Expression of Interest (EOI): Students are to submit their EOIs to Georgeta Bordea (name.surname@univ-lr.fr<mailto:name.surname@univ-lr.fr>) including the following documents: - CV including information about publications. - Transcripts of most relevant/recent degrees. - To apply, applicants must submit an EOI detailing their suitability for the project by addressing the required skills and key responsibilities in under 800 words. [1] https://l3i.univ-larochelle.fr/Images-et-contenus [2] https://l3i.univ-larochelle.fr/ [3] https://siren.io/ -- [https://d4d8xd20er8lg.cloudfront.net/3991be85-ba5e-4c61-894c-a2fd3bbbfbb4/rS...] Renaud Delbru Chief Scientific Officer & Co-Founder, Siren [https://cdn.gifo.wisestamp.com/social/rich-field-address-2/09095C/26/trans.p...] Avenue Roul 109, CS 40007, Talence, Bordeaux 33405, France [https://cdn.gifo.wisestamp.com/social/rich-field-phone-3/09095C/26/trans.png] +353 83 191 4812 [https://cdn.gifo.wisestamp.com/social/rich-field-email-1/09095C/26/trans.png] renaud.delbru@siren.io<mailto:renaud.delbru@siren.io> [https://cdn.gifo.wisestamp.com/social/rich-field-website-1/09095C/26/trans.p...] www.siren.io<https://www.siren.io/> [twitter]<https://twitter.com/sirensearch> [youtube]<https://www.youtube.com/channel/UCKGsC-vD28r7hW6T9QspKPA> [linkedin]<https://www.linkedin.com/company/11117365> [__tpx__]