Phd automatic Markers Detection By a

**About the role**:
Your role is to carry out a thesis work on: “Detection of markers by a psycholinguistic approach in small CR corpus”
- In the context of Customer Relations, listening to customers is a major focus of responsiveness and quality. In the mirror, listening to employees allows the implementation of a symmetry of attentions necessary for an efficient RC: expressive dimension of work, regulation (Dupuis, J., 2020).
Beyond the collection, the research team finds gaps in the field of verbatims processing in terms of speech analysis and highlighting the psychological state of speakers, despite the many existing tools to analyse “voice of customers”:

- tools that mainly shed light on a quantitative point of view and do not make it possible to understand the speaker’s experience (client or employee) that is expressed,
- approaches to the analysis of reducing human behaviors, by current algorithms, concerning the analysis of verbatims,
- poor consideration of context in analytical methods that fail to account for speaker experience.

**Problematic**: Can the use of theories in psychology and linguistics help improve speech analysis tools and results, especially corpus annotation and help to define categories, such as mental load, emotions, relationship to others or relationship to time?

**Scientific objective - results and locks to be removed**
The aim of the phd is, on the one hand, to characterize categories of psychological states, based on theories from psychology and linguistics. On the other hand, it is a question of defining an efficient method of annotation and of feeding our machine learning tools, to improve the tooling of speech analysis in the corpus of customer relations.

It will be necessary to successfully adapt existing models (especially in psychology and sociology) to our use cases, in order to create an efficient methodology of multidimensional qualitative analysis of the corpus of unstructured textual data, as well as intelligent, multi-dimensional (TAL & IA) analysis automation.

**About you**:
**Skills (scientific and technical) and personal suitability required by the position**

Strong knowledge in cognitive sciences (Linguistics, cognitive psychology, computer science) and automatic language processing. Curiosity for other disciplines, ability to adapt and integrate into a multidisciplinary and multi-site team, knowledge of the main TAL tools and knowledge of different computer languages (Python, etc.).

Very good writing skills.

**Training required (master’s, engineering degree, PhD, scientific and technical field, etc.)**

Master II research in psycho-linguistics, Master II in language sciences, Master II research in TAL, with training in cognitive sciences.

**Desired experiences (internships, etc.)**

Experience in TAL and using models in cognitive psychology.

At least one internship in a company.

**Additional information**:
This thesis is part of a particularly dynamic field and offers an original research angle: the approach to the treatment of a particular type of language, using models and knowledge, particularly in cognitive psychology. The type of language studied is a corpus of messages that are of variable size, obtained in the context of a particular asynchronous interaction, resulting from the use of a research tool at Orange Innovation.

**Department**:
Orange Innovation brings together the research and innovation activities and expertise of the Group's entities and countries. We work every day to ensure that Orange is recognized as an innovative operator by its customers and we create value for the Group and the Brand in each of our projects. With 720 researchers, thousands of marketers, developers, designers and data analysts, it is the expertise of our 6,000 employees that fuels this ambition every day.

Orange Innovation anticipates technological breakthroughs and supports the Group's countries and entities in making the best technological choices to meet the needs of our consumer and business customers.
In the Customer Relationship & Business Information System department, you will join the multidisciplinary research team of the Digital for hUman Program, led by Caroline Dubois. The team’s work uses the synergies between technologies (AI, TAL, etc.) and cognitive sciences (cognitive psychology, linguistics) to design and test tools for Customer Relations. This research ecosystem benefits from adapted infrastructures (test rooms, equipment) and data from customer relations. You will be asked to publish your work in academic papers and conferences.

**Contract**:
Thesis