Top

DEIM-16

This is the page which provides the source code and dataset we used for DEIM-16.

• Paper: Inducing Writers' Values on Concept Ordering from Microblog
• Authors: Tatsuya Iwanari, Naoki Yoshinaga, Masashi Toyoda, and Masaru Kitsuregawa.

Source

We made a front-end system (Ruby on Rails) and a back-end system (C++, Ruby, etc.). They are not refactored and will not be maintained.

• See a snapshot (not refactored) at GitHub.
  • We have a plan to maintain the code in an open repository (Under construction…).

Dataset

We make 2 types of dataset available on this page (all of them are written in Japanese).

humans (for gold-standard)

We gathered 100 Twitter users by crowd sourcing. This data contains the human annotations to make the gold standard ordering. They are written in Japanese.

Download:

• data

The data has 9 domains (3 for regions * 3 for genders);

• region
  • all
    • all users
  • kanto
    • users living in Kanto region
  • kinki
    • users living in Kinki region
• gender
  • all
    • all users
  • men
    • male users
  • women
    • female users

Please see human annotation format section for more detail. We also provide the original data and the annotation manual.

orderings (system-generated orderings)

This data contains the gold and system-generated orderings. They are written in Japanese.

Download:

• data (Japanese)

The data has results of 4 types methods for the 9 domains;

• types (please refer to the paper);
  • gold
  • base
  • adjs (+adjs)
  • gen (+gen)
  • both (+both)

Please see orderings format section for more detail.

Contact

If you have trouble to use them, please contact me (Tatsuya Iwanari nari@tkl.iis.u-tokyo.ac.jp).