SEntiment analysis and EMOtion technoLogies for social empathic Agents in ambient intelligence

Status

In progress

Proposed start date

2016-05-12

Proposed end date

2018-12-31

Project leaders

Carlos A. Iglesias; Tomás Robles

Description

Today, users are used to expressing their opinions and emotions in social networks. As usual, users make decisions based on the opinions of their circle of acquaintances, and they pass on the emotions shared. This has been of interest from the point of view of business to determine, for example, which users have the ability to influence others in order to increase the impact of advertising or political messages, or to improve the positioning of a brand. However, the field of emotions management is not limited to the virtual world of social networks, and is coming to our physical environment. Thanks to the devices based on ambient intelligence, we can adapt workspaces and residential areas to the users’ mood, and use it in malls to increase their purchasing.

The SEMOLA project aims to research on models, techniques and tools for the development of the empathetic personal agents endowed with a model of emotions. This facilitates the management of users’ relations with an intelligent social environment consisting of sensors and ambient intelligence. To this end, the project aims to: i) investigate recognition techniques and spread models for sentiments and emotions in social networks and smart environments (for that purpose, semantic web and linked data technologies, natural language processing, and social simulation will be employed) ; ii) research on customization services based on the context given by ambient intelligence devices in smart environments; and iii) research on agents models capable of generating emotions with a conversational interface, and to serve as a decision support system to assist users of an intelligent environment in their daily lives.

The SEMOLA project validates the results in two case studies: i) management of social networks as Twitter and Facebook and ii) an emotional and intelligent workplace. Finally, to ensure the robustness of the results, SEMOLA researches on verification and validation of case studies by agent-based social simulation models addressed in the project; such as the diffusion model of sentiments and emotions, or the infrastructure of sensors and actuators for ambient intelligence.

 

Project funded by the Spanish Ministry of Economy  (TEC2015-68284-R).

Publications View all