We are creating a tool to help you learn more about people around you, and connect with those with similar interests or background.

We are developing:

• a back-end tool to analyse public profile of a person
  • input:
    • person's Twitter and LinkedIn ID
    • (optionally: also Google Scholar, FB? or even person's for web search?)
    • (optionally: the category - professional, hobby, all?)
  • output: tags for that person (+ weights? likelihoods? categories?)
• a front-end (web) interface to the back-end tool

• Data collection – sources -> text

  • Azqa, Maria, Marija, Paweł, Mufutau
  • API (limitations) and library investigation
  • What data and metadata is available?

• Text processing – text (and metadata) -> tags

  • Sabrina, Konst., Marija, Azqa, Maria
  • how? What library to use (nltk, gate)?
  • dealing with synonyms, categories

• Information analysis – tags (words, cats, weight, time?) -> tag cloud

  • Sabrina, Sebastian, Paweł, Mufutau, Konst.
  • Testing different thresholds, weights, timing, person vs. the contacts
  • Building tag cloud (profile) of a person

• Front-end (web) interface

  • Harris
  • taking user id (Twitter handle, LinkedIn profile) as input
  • visualisation of results

• Linux virtual machine, login using SSH (ssh [your CERN login]@tedxapp on Linux or Putty on Windows)
• Firewall open for specific port:

sudo firewall-cmd --zone=public --permanent --add-port=8443/tcp
sudo systemctl restart firewalld

• http://www.nltk.org/install.html
• http://www.nltk.org/data.html
• https://github.com/nltk/nltk/wiki/Installing-Third-Party-Software
  • http://nlp.stanford.edu/software/lex-parser.shtml
  • http://nlp.stanford.edu/software/CRF-NER.shtml
  • http://nlp.stanford.edu/software/tagger.shtml
  • all ZIPs downloaded to /opt/nltk and unzipped
  • /etc/environment contains CLASSPATH=/opt/nltk/stanford-ner-2015-04-20 STANFORD_MODELS=/opt/nltk/stanford-ner-2015-04-20/classifiers
  • try python src/test/nltk-test.py

• https://docs.google.com/document/d/1vJncqSeDNEUWGPD0mNYWFLoSl7xSUsobCwbG88O2IJc/edit?usp=sharing
• the team: webfest2015-who-are-you@cern.ch (https://e-groups.cern.ch/e-groups/Egroup.do?egroupName=webfest2015-who-are-you -> members)
• initial proposal: https://webfest.web.cern.ch/content/who-you-are-analysing-public-linkedin-and-twitter-profiles