Personal Language Analytics for Emotion, Sentiment and Personality Modelling
Unit: MLDAT/PARSEM

Proposers : Caroline Brun & Scott Nowson
Duration: 4-6 months
Start Date: March 2016


Description

Personal Language Analytics (PLA) is an approach to text mining
whereby the focus is on the authors of texts rather than the texts
themselves. It is a computational field which combines aspects drawn
from natural language processing, data mining, linguistics, psychology
and sociology.

At XRCE we are interested in understanding how language can be used
across cultures to express mental states, like sentiment, emotion or
mood; to convey a sense of an individual's personality.  This
internship will explore the intersection of these areas as part of a
much larger project on customer modelling and personalisation.

We are particularly interested in the areas of:

- textual expression of emotion in human dialogue; 
- grounding and focus of this emotion;
- the relationship between the degree of emotional expression and
    personality traits of the interlocutors.

Tasks/Responsibilities

- Contribute to design of annotation schemes and corpora annotation
  for complex PLA tasks using dedicated annotation platforms.

- Work with and extend existing linguistic processing and text analytics tools

- Develop prototype text classification models and design experimental
  evaluation program.

Ideal candidates are Masters or PhD students with: a strong background
in natural language processing, experience in linguistics along with
text/data mining and machine learning; preferably an ability to script
(i.e. python); and ideally an interest in human language use.

Application instructions

Informal inquiries are welcome and can be made to
scott.nowson@xrce.xerox.com or caroline.brun@xrce.xerox.com .

To submit an application, please send your CV and cover letter to all
of: xrce-candidates@xrce.xerox.com ; caroline.brun@xrce.xerox.com and;
scott.nowson@xrce.xerox.com .