Datasets

The enriched graph used for join operator. The vertices have been enriched following the LDBC Social Network Benchmark protocol. The TXT files contain vertices and edges in a comma and tab separated format.

1. Enriched Graph Data [344MB]
2. LiveJournal edges

We also provide the dataset that we used to benchmark the graph nesting operator. In particular, we offer both the gMark -generated subgraphs and authorship's Microsoft Academic Graph subgraphs.

1. gMark-generated operands [2GB]
2. Microsoft Academic Authorship graph [5GB]

The Facebook posts of six public figures from different social categories (i.e. politicians, journalists and singers). Each JSON file includes one thousand posts.

1. Amanpour [205KB]
2. Macklemore [214KB]
3. Obama [136KB]
4. Renzi [569KB]
5. Travaglio [2MB]
6. Vasco [410KB]

The evaluation dataset is created using PubMed abstracts. Each entry of the dataset contains articles related to a disease.

1. Dataset creation procedure.

The dataset includes photos taken by 19 different smartphones, both from the front camera and the rear camera. For each smartphone a subset of 100 images (50 from the front camera and 50 from the rear one) was uploaded and downloaded on the following Social Media: Facebook, Flickr, Google+, GPhoto, Instagram, LinkedIn, Pinterest, QQ, Telegram, Tumblr, Twitter, Viber, VK, WeChat, WhatsApp and WordPress. The Readme.csv file summarizes the smartphones' characteristics.

1. Images [53GB]
2. Readme

The dataset includes videos taken by 13 different smartphones, both from the front camera and the rear camera. The Readme.csv file summarizes the smartphones' characteristics.

1. Videos [80GB]
2. Readme

This dataset comprises the time intervals extracted and normalized from the temporal expressions found in text corpora.

1. Wikipedia - 89 Million timexes [8GB]
2. New York Times 1987-2007 - 15 Million timexes [2GB]