This readme file describes (1) how to import the data into a MySQL database, (2) the schema in which the data are stored, and (3) the instructions given to human annotators when the data were collected.

* Importing

BaumerEtAl_NAACL-HLT-2015.sql contains the output from a mysqldump command and can be read into a MySQL database using the standard mysql restore process (mysql -u <user> -p<password> FrameAnnotation < BaumerEtAl_NAACL-HLT-2015.sql). This assumes you have already created a database with the name FrameAnnotation.

* Schema

The data are stored in two tables: Documents and Annotations.

Documents includes the following columns:
doc_id (unsigned integer) - A unique identifier for each document.
doc_html (text) - The text of the document with some simple title and paragraph HTML mark-up to ease reading for human annotators.

Annotations includes the following columns:
ann_id (unsigned integer) - A unique identifier for each annotation.
a_id (integer) - Identifier for the annotator who completed this annotation. Some annotators annotated more than one document, but not document was annotated by a single annotator more than once (integer).
doc_id - Identifier for the document to which this annotation applies. Corresponds with doc_id in the Documents table.
char_annotation (text) - Text field that stores the annotations. Each highlighted passage is denoted by the character index (zero-based) at which the annotations begins and at which it ends, with a space between the two indices. Annotations are separated by a period character. So, e.g., if char_annotation contains "78 90.173 180." then this annotator highlighted two passages, one from characters 78 to 90 (inclusive), and one from characters 173 to 180.
annotation (text) - Text containing a binary array where each digit represents one word in the document. A 1 indicates that the annotator highlighted the word, and a 0 indicates that the annotator did not highlight the word. This field is used primarily for convenience; char_annotation should be considered the definitive, unambiguous record of the annotations.
valid (tiny int) - Indicates whether the annotation passes the criteria for a valid annotation described in the Baumer et al. NAACL-HLT 2015 paper in which these data were used. 1 means the data are considered valid, 0 means they are not.

* Prompt

Below is the prompt that was given to human annotators who completed the annotation task.

---

Political issues can often be complex, contentious, and difficult to understand. One way of making sense of these issues, and the different positions that one can take on an issue, is to think about the frames that structure debate about the issue. Frames help organize facts and information. They help define what counts as a problem, diagnose the problem's causes, and suggest remedies for solving the problem. These ways of thinking have lots of different parts, including stereotypes, metaphors, images, catchphrases, and other elements.

These different frames are often associated with a particular way of talking about or communicating about an issue. Certain words or phrases might suggest that one or another frame.

Please use the tool at the link below to highlight the words or phrases that help you identify the framings used in the article you just read.

---