Built at Universitat Jaume I for the GRAPE research group
GRAPE MARS is a video annotation and analysis platform for researchers who study how people talk and act together. You mark time ranges on tracks, transcribe the speech, and turn the marks into charts and exports.
Runs in the browser. Nothing to install.
What you can do
Annotate
Each track holds one kind of annotation. Drag across the timeline to mark a time range, and the mark takes the color of its type.
Add speakers and code one person at a time. Count occurrences on some tracks and measure duration on others.
Transcribe and tag
Transcription runs on the site and costs you nothing. Click a word and the video jumps to that moment.
State the language of a video and GRAPE MARS tags each word with its lemma and part of speech. It works in Spanish, Catalan and English.
Analyze
A workspace is your own page of modules: distributions, co-occurrence, concordance, inter-rater agreement and more.
Set the scope once and every chart follows. Export a workspace to Excel, or send your annotations to R or ELAN.
Work together
Invite colleagues to a project and give each one a role. Each annotation records who made it, so two coders can work on the same video.
Then compare them with an agreement score before you divide the rest of the corpus.
How it works
The help center walks a new account to an exported result in twelve short pages. These are the four stops that matter most.
A project is a folder for one study. Add the videos it covers and state the language they are in.
Read the guideDecide which types you will code, put each on a track, then drag along the timeline to mark them.
Read the guideTurn the speech into text, name who is talking, and let the tagger read the transcript word by word.
Read the guideAdd the modules you need, read the charts, and export to Excel, R, ELAN or a Word report.
Read the guideHelp center
One task on each page. Every error message the app can show, with its meaning. Every word the interface uses, explained.
Made at a university, for research
GRAPE MARS is built at Universitat Jaume I for GRAPE, the Group for Research on Academic and Professional English. The group studies discourse and multimodal communication in academic and professional settings.
The group uses the tool in that research. It codes gesture, gaze and talk in recorded lessons, then counts and compares across a corpus. The tool replaces an earlier GRAPE MARS, and it imports projects from that version.
Other groups who work with video are welcome to try it, and to tell us what they need.