Dataset Hub

The collection comprises multiple multi‑view benchmark datasets for clustering or classification tasks, including AwA, COIL100, MNIST2, NUSWIDEOBJ, PIE, and YouTubeFaces, each accompanied by detailed descriptions and multi‑view features.

The dataset is named danbooru tags and mainly includes the names of Danbooru tags, the words that compose tag names, and the tag categories. Tag categories include General (0), Artist (1), Copyright (3), Character (4), and Metadata (5), etc.

This dataset consists of a series of email messages and serves as training and testing data. Because the dataset is used for a competition, the test set is unlabeled while only the training set is labeled. The training set contains 2,929 legitimate (ham) emails and 1,378 spam emails.