Please make sure to follow the rules:
- Submission: 3 times a day, interval of 1hour between submissions
FASCODE (FAShion COordination DatasEt / FAShion CODE) is a dataset that contains conversations between an AI fashion coordinator and users for recommending outfits. It includes information such as speaker attributes, conversation order, tagging information for conversation functions, and text information regarding the recommended clothing. (Refer to the data and codebook for detailed information.)
※ The dataset provides two versions: one with token segmentation (word segmentation) applied and one without.
※ Please check the frequently asked questions in the 'Notice and FAQ' bulletin board!
The 'Notice and FAQ' bulletin board is a participant support page created with Notion and can be accessed on Chrome, Edge, Mobile, Safari, etc. It does not run on Internet Explorer.
Evaluation Metric: WKT
WKT stands for Weighted Kendall's tau. Kendall's tau is a type of rank correlation coefficient that compares the rankings between two variables to measure their association. Weighted Kendall's tau is a method of measuring the correlation coefficient with weights taken into account based on the rankings.
Develop an evaluation system/artificial intelligence during the competition period using the training set and development set of FASCODE to learn a method of recommending outfits based on user speech and selecting appropriate fashion coordination. Develop and compose a model, then submit it to the leaderboard.
- Training set: FASCODE dataset
* Refer to the specific Sub-Task limitations (see the table above).
- Development set (Evaluation set) released (simultaneously with the training set)
- Provided data input/output and evaluation codes
- FASCODE encoding: EUC-KR
* Submission: 3 times a day, interval of 1hour between submissions
※ By submitting the model, it is assumed that you agree to the calculation of the test score and the public disclosure of the score through the leaderboard.