The May’s PGRs Research Presentations was held on Wed. 14th June, 11am, Meeting Room, MC3108 (3rd floor).
This session we had the following presentations:
| Title: “Automatic Analysis of the Social Behaviours of Fish using Computer Vision“. | Title: “Investigating text analysis of user-generated contents for health related applications” | |
|
By: Alaa Al-Zoubi |
By: Deema Abdal Hafeth |
|
| Abstract: The development of computer vision as a method for automatically monitoring and analysing human activities is a well established research area. However, the application of this technology to support analysis of animal behaviour is a relatively new area of research that is attracting increasing attention from both the computer science and biological science communities. Current state-of-the-art fish monitoring systems are lack of intelligent in interpreting fish behaviors automatically. Sticklebacks have been a model species in behavioral biology for over half a century. Traditional methods of studying the social behavior of these fish involve manual observation and recording. However, these methods are time-consuming, potentially error prone, and a limiting factor on the amount of data which can be analysed. To tackle these problems, We are developing a computer vision system to automatically detect and track the social behaviors of sticklebacks, under laboratory conditions. Our system will provide automated quantitative measurements for researchers to collect and analysis stickleback’s behaviors. The system will have the ability to deal with large dataset for a long period of time to facilitate studying the sticklebacks life cycle. |
Abstract: Data in patients’ records includes free-form text, which have valuable medical related information embedded in. This data can be extremely useful in aiding and providing better patient care. Text analysis techniques have demonstrated the potential to unlock such information from text. One challenge with clinical reports’ data is their strict availability and difficulties in accessing them. On the other hand, people are expressing themselves more widely nowadays and the online user-generated contents (UGC), like forums and blogs, are becoming more available. The aim of this work is to investigate the potential of text analysis techniques in predicting the smoking status but from user-generated contents such as forums. This especially includes the use of Psycholinguistic features on analysing forums, with the hypothesis that forum posts have different linguistic features and are rich in personal stories, fresh opinions, and thoughts. | |
The Q/A was followed by a brief cath-up meeting.