Finding Influential Institutions in Bibliographic Information Network

• This is the code used in KDD Cup 2016 solution "Finding Influential Institutions in Bibliographic Information Networks".
• This solution achieved 11th position in the contest. Full paper describing the solution is available at https://arxiv.org/abs/1612.08644

Instructions to run the code

• Download following files in directory data:

• Following columns should be present in each file

  SelectedAffiliations.txt

  Col #	Col Name
  1	Affiliation ID
  2	Affiliation name

  Authors.txt

  Col #	Col Name
  1	Author ID
  2	Author name

  Conferences.txt

  Col #	Col Name
  1	Conference series ID
  2	Short name (abbreviation)
  3	Full name

  ConferenceInstances.txt

  Col #	Col Name
  1	Conference series ID
  2	Conference instance ID
  3	Short name (abbreviation)
  4	Full name
  5	Location
  6	Official conference URL
  7	Conference start date
  8	Conference end date
  9	Conference abstract registration date
  10	Conference submission deadline date
  11	Conference notification due date
  12	Conference final version due date

  FieldOfStudyHierarchy.txt

  Col #	Col Name
  1	Child field of study ID
  2	Child field of study level
  3	Parent field of study ID
  4	Parent field of study level
  5	Confidence

  SelectedPapers.txt

  Col #	Col Name
  1	Paper ID
  2	Original paper title
  3	Normalized paper title
  4	Paper publish year
  5	Paper publish date
  6	Paper Document Object Identifier (DOI)
  7	Original venue name
  8	Normalized venue name
  9	Journal ID mapped to venue name
  10	Conference series ID mapped to venue name
  11	Paper rank

  PaperAuthorAffiliations.txt

  Col #	Col Name
  1	Paper ID
  2	Author ID
  3	Affiliation ID
  4	Original affiliation name
  5	Normalized affiliation name
  6	Author sequence number

  PaperKeywords.txt

  Col #	Col Name
  1	Paper ID
  2	Keyword name
  3	Field of study ID mapped to keyword

  PaperReferences.txt

  Col #	Col Name
  1	Paper ID
  2	Paper reference ID

• The code is available in the directory named "Code".

• All the code is written in Python 2.7

• To run the code, go to the directory "Code" and run the file "main.py".

Author

• Anubhav Gupta