**Apologies for cross-postings**

Applications are invited for a 12-month-postdoctoral fellowship in
Computer Science. The research will be conducted in the DM2L team of the
LIRIS lab (CNRS UMR 5205) at INSA Lyon, Villeurbanne, France. The
position is funded by IDEX Lyon IMPULSION.

Project information

Context

With the emergence of artificial intelligence, the extraction of
geographic information has become a key element in knowledge extraction
and information retrieval. It involves multiple fields especially
related to digital humanities, such as information processing, knowledge
management, automatic data analysis and decision making. It is
increasingly attracting the attention of many researchers to go beyond
the possibilities offered by traditional geographic information systems
to manipulate named places in addition to geographical coordinates. In a
context of management and analysis of big data, the existing methods
appear too segmented and too specific to the type of documents
considered. The generalization of a method for the treatment of
heterogeneous data then appears as a major scientific issue.

One of the objectives of this project is to strengthen interdisciplinary
collaborations and federate research in engineering (computer science
and data science) and Humanities and Social Sciences (geography,
history, linguistics, etc.) around digital issues. The challenge is at
the same time to solve scientific issues in engineering and to address
societal issues around the treatment of geographic information (smart
city, heritage enhancement, territory development, etc.).

The post-doctoral candidate will integrate with the DM2L team of the
LIRIS lab and their research objectives will be related to several
ongoing projects (the CHOUCAS ANR project, the GéoDisco MSH Lyon
St-Etienne project, and the Spatial History of the Encyclopédie
project). The candidate will also have the opportunity to participate in
multidisciplinary international collaborations (with computer
scientists, geographers, historians, and linguists).

Research objective

In this project, we propose to design and develop a hybrid and modular
method of extracting geographic and semantic information from
heterogeneous textual corpora (eg, historical, legal, literary, tourist
documents, user reviews, social media, etc.). In this context there are
several research challenges to be addressed. The first objective is the
identification and automatic classification of documents in order to
apply the most appropriate information extraction methods. Then the main
objective concerns the combination and interoperability between symbolic
and statistical NLP (Natural Language Processing) methods (symbolic
rules, machine learning, and data mining). This combination of
approaches should lead to better performance than conventional methods
and better adaptability to the type of document. In particular, we want
to develop and implement a hybrid method based on formal grammars and
supervised learning techniques.

The overall objective of the project is to design a platform (which will
be deployed as web services) to combine, enrich and evaluate the latest
advances in the field of NLP, semantic modeling and knowledge
extraction. The different methods implemented within the platform will
be accessible to non-experts.

 
Practical information

Knowledge, Skills and Experience

- Ph.D. in Computer Science, Data Science, Artificial
  Intelligence/Machine Learning, or a related field.
- Strong publication record and/or proven ability to deliver
  high-quality software in a research context.
- Good communication skills (oral and written).
- French is a plus but not mandatory.

Remuneration terms and Conditions

- The starting date is flexible, between February and November 2019. The
  candidate should indicate the preferred starting date in the
  application letter.
- The position is for one year (12 months) with a salary after taxes of
  2,100 euros per month (gross monthly salary: 2,600 euros).

Application procedure

- Enquiries should be directed to Dr Ludovic Moncla
  (ludovic.moncla_AT_liris.cnrs.fr).
- Applications including a CV, a list of publications and an
  approximately two-page description of research interests should be
  sent by email to Ludovic Moncla: Applicants should also arrange two
  recommendation letters (to be sent to the same address).
- Applications will be welcome until the position is filled, those
  submitted before 15/02/2019 will be considered in priority.