Milestone 5

KL metric (Marc):

• cross-validation for dating articles:
  for one year:
      for(int i = 0, i < 100, i++) {
          randomly select a subset of articles
          date it
          compute the estimation error
      }
      return the average of the errors obtained
• Test the metric with n-grams for n>1

Out of place metric (Tao):

• Debug
• Test with and without ocr correction
• Test with n-grams for n>1
• Date a subset of articles

Other metrics (Cynthia):

• Debug articles dating
• Test with different weighting (ex tf-idf)
• Test with n-grams for n>1

Synonyms (Nicolas):

• Implement article dating using synonyms

Topics clustering (Farah, Jeremy):

• Debug topic clustering and create a directory for each topic containing the 1-grams of the corresponding articles
• Apply one or more metric on each topic separately

Sentence analysis (Gil, Malik):

• Implemet article dating using your analysis