README.md

CIKMAnalytiCup-2020 COVID-19 Retweet Prediction Challenge

overview

usage

download data (TweetsCOV19)

• register codelab and join "COVID-19 Retweet Prediction Challenge"
  • https://competitions.codalab.org/competitions/25276
• download data (https://competitions.codalab.org/competitions/25276#participate-get_starting_kit)

run

$ cd ./run
$ sh solution.sh

results

• 3rd palce / 200+ participants
• Final leaderboard of the top six teams (semi-finalists)
  • competition page: https://competitions.codalab.org/competitions/25276#results

Rank	Team	MLSE
1	vinayaka	0.120551
2	mc-aida	0.121094
3	myaunraitau (ours)	0.136239
4	parklize	0.149997
5	JimmyChang	0.156876
6	Thomary	0.169047

folder structure

CIKMAnalytiCup-2020

├── exp # deploy the experiment's config
     └── (experiment name)
             └── config.py

├── features # save extracted feature files
     └── (feature_name).feather

├── folds # save cross validation data table files
     └── (folds_name).csv

├── input
     └── public_dat
         └── feature.name
         └── train.data
         └── validation.data
         └── test.data

├── run
     └── solution.sh
     └── ensemble.py

├── save
     └── model/
     └── model_predict/
     └── predict/

├── src
     └── create_folds.py
     └── dataset.py
     └── feature_extraction.py
     └── model.py
     └── train_5fold.py
     └── utils.py
     └── utils_feature.py
     └── utils_model.py

├── submission # save submission files
     └── (submission_name).zip

Contact

• Email: daichi.takehara0730@gmail.com (Daichi Takehara)