CCU_Rater_Onboarding

Here at CCU Editorial Tagging, we're dedicated to ensuring our content is accurately categorized. This page serves as your comprehensive guide for onboarding, outlining the rationale behind our work, the tools and documents you'll utilize, and how to effectively rate tags. We also provide real examples of common mistakes to help you avoid them. The onboarding process begins with three clear steps. First, we encourage you to read the Training tab thoroughly from beginning to end. Once completed, keep it open as a readily available reference. Second, you'll engage with practice quizzes. These consist of sets of approximately 10 questions. After each answer, you'll see the correct response and a detailed explanation. You can generate new sets of tags as many times as needed until you feel fully prepared. Finally, you'll take the real test, which comprises 20 tags with no answers shown. It's crucial to take and submit this test only when you're confident in your understanding. Passing this real test is a prerequisite for consideration for future work on our datasets. Therefore, we emphasize the importance of completing the practice quizzes first to build your proficiency before attempting the final assessment. Our raters play a crucial role in maintaining the quality and relevance of our content catalog. Every task involves a title and a list of suggested tags. Your responsibility is to determine which tags are appropriate for the title and the degree of their relevance, all while adhering to the guidelines provided in this guide. The ultimate goal is to apply our guidelines effectively and prevent common errors. A key principle to remember is that a tag's definition might seem correct but could still be unsuitable for a given title. Before approving any tag, ask yourself two critical questions: Would a member browsing this tag be happy to discover this title, and would this title fit naturally alongside the example titles listed in Graffiti? These two tests are fundamental and are covered in depth within our documentation. Every aspect of this guide is built upon these two foundational questions. We believe that understanding "Why This Work Exists" is essential. The bigger picture involves editorial tags, which describe the core characteristics of a title. These tags encompass genre, subgenre, form, storylines, tones, and episode types, collectively painting a comprehensive picture of what a title truly represents. The process unfolds in several stages: a title arrives, suggestions for tags are generated by a model, and then you, as the rater, evaluate these tags. Your ratings are crucial because they teach our model to suggest even better tags in the future. Accurate and consistent ratings ensure that our system continuously improves, leading to a more refined and precise tagging process for all our content. This iterative feedback loop is vital for the ongoing development and optimization of our editorial tagging system.