The collection of blog posts with sentiment markup

Description

This collection is consists of blog posts, which were used in testing ROMIP-2011 sentiment classificaiton tasks.
Each post is from one of three domians (books, movies or digital cameras) and has its sentiment score on two, three or five point scale.
Additionaly in each post all objects (main and secondary) from the corresponding domain are indicated.

Dataset Parameters

• Collection size: 2,5 Mb
• Quotes number: 874
• Encoding: windows-1251

Rights to Use

To get access to the collection you must sign the usage agreement.

Data Format

The collection is distributed in xml file of a certain format.