Take streaming tweets, extract hashtags & usernames, create graph, export graphml for Gephi visualisation
Write-ups:
- http://ianozsvald.com/2013/03/22/analysing-pydata-london-and-brighton-tweets-for-concept-mapping/
- http://ianozsvald.com/2013/03/18/semantic-map-of-pycon2013-twitter-topics/
The code here is based on Maksim Tsvetovat's tutorial at PyCon 2013 https://us.pycon.org/2013/schedule/presentation/29/
Turn a set of streamed Twitter firehose JSON tweets into a Twitter concept net showing commonly discussed pairs of hashtags and users.
Here we see a plot from my blog post http://ianozsvald.com/2013/03/18/semantic-map-of-pycon2013-twitter-topics/ demonstrating the hashtags and usernames that were grouped together at PyCon 2013. White is #hashtags, purple is @usernames:
Using the above data inside Gephi I extract a set of communities:
You need a set of streamed tweets, I can't provide my set as the Twitter Usage Agreement forbids distribution of raw tweets. I've included a summary file which can be used in networkx (and gephi) which does not include tweet text or creation dates (as per the Twitter requirement).
You can get your own set of streamed tweets using:
$ curl -s -uUSERNAME:PASSWORD -X POST -d 'track=pycon' https://stream.twitter.com/1.1/statuses/filter.json -o tweets_pycon.json
You can leave this running via a script (e.g. get_tweets.sh) using:
nohup curl -s -uUSERNAME:PASSWORD -X POST -d 'track=pycon' https://stream.twitter.com/1.1/statuses/filter.json -o tweets_pycon.json > nohuppycon.log &
You also need the following:
* twitter-text-python
* networkx
The easy way to get these is to install:
$ pip install -r requirements.txt
$ pip install -r requirements_2.txt # installs matplotlib after numpy from first requirements file
Note the above use of 2 requirements files is due to a matplotlib setup issue: https://github.com/matplotlib/matplotlib/wiki/MEP11
To parse the raw Twitter JSON into a cleaned local version use:
$ extractor_content.py --json-raw eg_tweets_pycon.json -o clean_pycon.json
Note that you have to provide eg_tweets_pycon.json, I do provide a clean_pyson.json from my data.
To build a graph network from the cleaned data use:
$ extractor_content.py --json-cleaned clean_pycon.json --remove-nodes #pycon #python #pycon2013 @pycon
To build the above and draw the graph use --draw-networkx:
$ extractor_content.py --json-cleaned clean_pycon.json --remove-nodes #pycon #python #pycon2013 @pycon --draw-networkx
To save a graphml output (for importing into Gephi) use:
$ extractor_content.py --json-cleaned clean_pycon.json --remove-nodes #pycon #python #pycon2013 @pycon --write-graphml pyconout.graphml
Other arguments:
--remove_usernames_below n # strip users with less than n occurrences
--remove_hashtags_below n # strip hashtags with less than n occurrences
Output just the text from tweets in the named period:
$ python extract_tweet_updates_to_file.py --json-raw <something>.json --text-file text.txt --ff 2013-04-11T22:00:10.044491+00:00 --ft 2013-04-11T23:00:00+00:00
To output just a filtered set of data (keeping anything except RTs in the date range):
$ extract_tweet_updates_to_file.py --json-raw <something>.json --output reduced_raw.json --ff 2013-04-11T22:00:10.044491+00:00 --ft 2013-04-11T23:00:00+00:00
MIT
Copyright (c) 2013 Ian Ozsvald.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
Copyright (c) 2013 Ian Ozsvald
Remove various types of quotes to normalise phrases:
- how we all lost\u2019 (double quotes)
- how we all lost' (single quote)
- consider adding root URLs (e.g. bbc.co.uk) or URL titles
Perhaps following URL link shortners and then take the domain name as a new link in the graph (e.g. bbc.co.uk, techcrunch.com).