Twittener : topic modelling
Twitter is a popular social networking site which allows users to get information such as news and trends. However, Twitter being a text-based social networking site, may not be suitable for certain pockets of people such as the elderly, people who often multi-task and the less literate. As such, Tw...
محفوظ في:
| المؤلف الرئيسي: | |
|---|---|
| مؤلفون آخرون: | |
| التنسيق: | Final Year Project |
| اللغة: | English |
| منشور في: |
2017
|
| الموضوعات: | |
| الوصول للمادة أونلاين: | http://hdl.handle.net/10356/70218 |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
| الملخص: | Twitter is a popular social networking site which allows users to get information such as news and trends. However, Twitter being a text-based social networking site, may not be suitable for certain pockets of people such as the elderly, people who often multi-task and the less literate. As such, Twittener is an alternative for users to interact with Twitter. It allows users to listen to tweets, instead of the traditional way of reading them. This project aims to enhance the Topic Processor component of Twittener and introduce a trending algorithm for the Trend Detector component. The Topic Processor component generates the topics from the tweets crawled from Twitter using the combination of Latent Dirichlet Allocation (LDA) and SumBasic algorithm. The Trend Detector aims to generate trending topics within a particular time frame. The purpose of this report is to document the development and implementation of the enhancement to the Twittener system. |
|---|