Investigate social dynamics of memes based on Sina Weibo
Python JavaScript Shell CSS
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.
Time-based data.ipynb

Mitras : Mining memes

Mitras is a set of scripts used to detect, analyse and visualize memes the Chinese microblog Sina Weibo. You will need the Weiboscope dataset prepared by HKU It has been developed by Clément Renaud for his phD research.

Mining workflow :

The different workflows used in this research are documented in iPython notebook in the /doc folder.

  • es_* : plain-text search and mining with ElasticSearch and Kibana
  • hashtags_* : build and analyze a corpus of all hashtags in the datasets
  • pm_* : meme detection clustering algorithm using protomemes (Ferrara, 2013)

For visualisation, we use Matplotlib, d3js, Networkx and Gephi.

Data : the Weiboscope corpus

To create this project we will use the data provided by the project Weiboscope from HKU University, JMSC - link. The dataset contains sample data from 52 weeks of 2012 from more than 350,000 Chinese microbloggers who have more than 1,000 followers (Fu, Chan, Chau, 2013 ; Fu, Chau, 2013).

Note : this data has been anonymized

Data Set Statistics:

* Number of weibo messages: 226841122
* Number of deleted messages: 10865955
* Number of censored ('Permission Denied') messages: 86083
* Number of unique weibo users: 14387628
* 57 files, 18G

Download & Prepare Data

# to download the data
bash bin/

#  Downloaded: 57 files, 18G in 6h 42m 3s (803 KB/s)

# move the files to the data folder
mv data/datazip
rm -R

# remove zip files
ls data/datazip/*zip | xargs -i rm {}