Credibility analysis of tweets automatically is a hot topic in computer science. Existing approaches include comparing relative information from trusted media, evaluating user credibility by analyzing user graph propagation, applying machine learning methods to judge the credibility of giving tweets, etc. In this project, we are interested in how we can include domain knowledge in the task of credibility analysis of Chinese tweet. For example, if we are given a set of tweets about Bioinformatics, how can we judge the credibility of the tweets base on training data and some existing Bioinformatics information?
Supervisor: Dr. Lucas Hui
Student: Bill Yue
Inital Stage:
Textarea on a webpage
Final Stage:
Twitter API
Other options:
Web crawling/other social media
Natural Language Processing:
To derive meaning from human or natural language input
Segmentation:
To separate each word phrase
Statistical analysis
Sentiment analysis
Data Visualization
Display the result on a webpage.
Here is the progress of my project.
This shows the deliverables for the project requirements.
The final product of this project would be a website with serveral servers.
Here is my project schedule
1 September 2015 - 4 October 2015
11-15 January 2016
15-24 January 2016
17 April 2016
18-22 April 2016
3 May 2016
Click on an image to read the document
Dr. Lucas Hui
Bill Yue