Reading Emotions

Frequently Asked Questions

1. How can I learn more before I start classifying?
Check out our tutorial! When you first click on the workflow, a tutorial will display on the screen. You can click through this before starting. You can always refer back to the tutorial by clicking on the Tutorial tab to the right of the Task tab within the workflow. You can also check out our Field Guide of ambiguous cases for more guidance. It is located to the right of the task.

2. What do I do if I'm not sure of the answer?
Please give your best answer. We want to know what you think!

3. Can I write multiple emotions?
Yes! Just separate them by commas. Don't worry about capitalization. But spelling does matter!

4. Where do the passages come from?
The passages for this workflow are excerpted from books in this contemporary literature dataset. Interested in what book a single subject is from? If you click the information button below any given subject (the small 'i' in the black circle), you can find the book author and name from the file path. Maybe a few sentences can hook you enough to read the full book!

5. Can I see how many passages I have classified?
Yes, you can!

• Click on your user name in the upper right hand corner and select "My Profile".
• Select the tab “Your Stats”. This will show you all of the classifications that you have done for The Lives of Literary Characters and any other Zooniverse projects that you have participated in.

6. How can I ask a question that is not in this FAQ?
You can ask questions on the project's Talk page in the General forum. "Ask questions about specific passages" is for questions about specific passages. "Ask the researchers questions about the project" is for bigger questions related to the research project.

7. What's the data going to be used for?
We get this question a lot! This project is about understanding how stories are told and what they mean. This is a question scholars have been asking for over two-thousand years. Annotating large numbers of passages allows us to do two things: a) learn what many readers think about storytelling and b) train AI to reproduce your judgments at even larger scale.

This work supports one of the biggest innovations in the study of storytelling to occur in the past few years: we no longer need to rely on the views of individual scholars reading a tiny portion of available stories. Instead we can aggregate your views and use those to understand stories at large scale (thousands to millions of examples). This dramatically alters how we understand stories, from point of view (citizen readers not single scholars) to scale (no longer just a small canon of works).

So when we say we are going to train AI models this concretely means we use your annotations to teach Large Language Models how to imitate your judgments on passages the models haven't seen before. We are not in the business of making chatbots or robot writers to replace aspiring writers. Quite to the contrary! We want to understand what writers all over the world are doing so we don't have to rely on small, manually curated samples. This project is about foregrounding readers and writers not replacing them.

8. Where will the data go?
The data will be placed in the following repository dedicated to this project. It will be publicly available once we have cleaned and processed it. This takes a long time, but we hope for a 6-12 month turnaround for each project.