Real-time sentiment analysis of video calls
Loading...
URL
Journal Title
Journal ISSN
Volume Title
Perustieteiden korkeakoulu |
Master's thesis
Unless otherwise stated, all rights belong to the author. You may download, display and print this publication for Your own personal use. Commercial use is prohibited.
Authors
Date
2019-05-06
Department
Major/Subject
Computer Science
Mcode
SCI3042
Degree programme
Master’s Programme in Computer, Communication and Information Sciences
Language
en
Pages
53 + 5
Series
Abstract
In recent years, with ever-increasing internet connection speed and bandwidth, video-focused software has become more and more popular for both work and pleasure. Examples of such applications include Skype, BlueJeans or iOS Face- Time. These applications, and the various interactions facilitated by them contain lots of interesting data that we feel would be very fruitful to gather and analyze. Within the context of this thesis, we focused on evaluating the potential of collecting sentiment analytics from video teleconferencing both on an individual and group level, for the purpose of helping people reflect on their own behavior and regulate their emotions . To achieve this, we developed a composable, scalable microservice-based analytics pipeline for video and speech, and a browser-based web application to demonstrate it. We evaluated already existing solutions for gathering sentiment analytics, and integrated two of them into our analytics pipeline. The whole system was deployed in a virtualized container environment using Docker. Be- sides the pipeline and web application, we also designed and implemented some visualizations for the data that we gathered. In the end we developed a working prototype, although deeper analysis and evaluation of the actual accuracy of its results needs to be performed. Human emotions are rather difficult to quantize. We found that the current APIs and libraries publicly available for performing sentiment analysis are already quite accurate and feature-rich, and we expect them to get even better.Description
Supervisor
Korhonen, AriThesis advisor
Leinonen, TeemuKeywords
programming, web development, learning analytics, sentiment analysis, distributed systems, analytics