zinny-surveys 0.1.8

Surveys and specifications for rating Motion Picture titles. Used by Zinny, the movei rating app.

Zinny Surveys: Structured Media Evaluation

What's the skinny on the ciné?

Zinny is a tool for rating media, and the surveys here provide for a meaningful evaluation of consistent criteria across many titles and perspectives.

A collection of survey definitions

This repo is a collection of surveys used by the zinny-api and front end zinny-webui. They are published separately to allow for independent use, and to encourage community contributions.

Installation

see SETUP.md

Motivation:

Comparing movies across genres, styles, or time periods can be challenging. Surveys offer a structured approach to evaluate films by breaking down complex elements into measurable components. This approach captures the expertise behind a movie more comprehensively than a single score.

Surveys define measurable criteria, while weights establish their relative importance, making evaluations both meaningful and customizable.

Goals

• Standardization: Standardization: Provide curated surveys and weights for consistent and reliable evaluations.

• Customization: Easily create or modify JSON-based surveys and weights to fit specific needs.

• Community Contributions: Contributions are welcome to expand criteria or add new surveys. Submit pull requests or reach out to get involved.

Definitions

Surveys

Definitions

Surveys

Survey: A predefined set of criteria for evaluating a media title. Each survey includes:

• Metadata: Describes the survey's purpose, version, author, and related details.
• Criteria: Individual measures with attributes such as range and descriptions.

Criteria:

• Criteria are measurable attributes within a survey, defined as:
  • ID: A unique identifier (e.g., "artistry").
  • Name: A human-readable label (e.g., "Artistry").
  • Description: Details what the criterion measures.
  • Range: (optional) Specifies valid scoring values (e.g., [1,10]).
  • Value Labels: (optional) Explains specific values within the range (e.g., 0: "No quality", 10: "best quality").

Repository Structure

Repository Structure

surveys/
├── shared/                     # Institutionally approved surveys
│   ├── vfx.json
│   ├── picture.json
│   └── ...
├── community/                    # Community-contributed surveys
└── local/                        # reserved for local surveys

Example JSON Representation:

see Examples.md

Weight Presets

Weights: Weights determine the relative importance of each survey criterion. They are linked to specific surveys (and optionally, survey versions) and allow for multiple perspectives by using different weight configurations.

Repository Structure

weights/                          # Weighting configurations
├── shared/                     # Institutionally approved weights
│   ├── vfx_even_weights.json
│   ├── picture_even_weights.json
│   ├── picture_storyteller.json
│   └── picture_technologist.json
├── community/                    # Community-contributed weights
└── local/                        # reserved for local weights

Collections

Collection: Collections group related items, such as titles or surveys, into curated sets. Each collection includes:

• ID: A unique identifier (e.g., "favorites_2024").
• Name: A human-readable name (e.g., "My Favorites").
• Description: Optional context for the collection (e.g., "Movies to watch this year.").
• Items: A list of either titles or surveys in the collection.
  • Titles: Include name (e.g., "Madame Web") and year (e.g., 2024).
  • Surveys: Include name (e.g., "VFX Evaluation") and version (e.g., "1.0").

Contributing

We welcome contributions! If you'd like to report an issue, suggest a feature, or contribute code, please check out the CONTRIBUTING.md file for guidelines.

Acknowledgements

• Development was sponsored by Teus Media.