Skip to content
Permalink
master
Go to file
 
 
Cannot retrieve contributors at this time
53 lines (39 sloc) 1.74 KB

Institution data cleansing and name resolution

We use Microsoft Academic Graph for cleaning institution names and aliases.

The result of the process described below are stored in inst_fullname.csv (containing names, urls and wikipedia urls), and inst_alias.csv (containing mappings of multiple surfaces strings to a single institution as described in item 1 below).

1) Searching Aliases (using MAG interpret api)

e.g. (Searching name for a key)

aalto university --> aalto university
aalto university school of business --> aalto university
aalto university school of electrical engineering --> aalto university
aalto university school of science --> aalto university

We found 6231 keys from 6245 institution names (Updated 11/July/2019)

type count
name == key (primary name) 6231
name != key (alias name) 14
unregistered name 0

Common Error Types

  1. Different key values for aliases

e.g.

at t --> at t
at t labs --> at t labs

auburn university --> auburn university
auburn university at montgomery --> auburn university at montgomery

current solution: rely on MAG result as it is.

  1. Alias not registered or unrecognisable

e.g.

No result: department of systems biology
No result: school of industrial technology

current solution: use the name as a key

2) Finding Fullname and URL of the institution (using MAG raw data)

Affiliations.txt has (key, fullname, grid, url, wikipedia_url) tuples.

$ cut -f3,4,5,6,7 Affiliations.txt > inst_fullname
You can’t perform that action at this time.