DataForGood-2022: Difference between revisions
(Added link to launch presentation) |
(tweaks) |
||
| Line 1: | Line 1: | ||
=== Summary === | === Summary === | ||
[https://dataforgood.fr/ DataForGood] is | [https://dataforgood.fr/ DataForGood] is a French "association" that mobilizes tech to help citizen projects. We are part of [https://dataforgood.fr/projects/tags/saison-10 season #10], this is our 4th season. | ||
Main focus this year is on using machine learning to automatically categorize products so as to enable computing Nutri-Score and Éco-Score. | Main focus this year is on using machine learning to automatically categorize products so as to enable computing Nutri-Score and Éco-Score. | ||
| Line 9: | Line 9: | ||
'''Expected outcomes''': Have a new machine learning model, ready to deploy in production: | '''Expected outcomes''': Have a new machine learning model, ready to deploy in production: | ||
* That | * That uses more features to predict category: OCR data, nutritional information, eventually images or any feature. | ||
* Precision should be very high to be able to apply category automatically but the model should use | * Precision should be very high to be able to apply category automatically but the model should use its confidence to ask the user whenever needed. | ||
'''Impact:''' | '''Impact:''' massively increase the number of Eco-Scores and Nutri-Scores we are able to provide on the Open Food Facts platform. | ||
Categorization is also important in many | Categorization is also important in many ways, for example to compare nutrition data with other products of same category (rank within a category). | ||
'''Timeline''': | '''Timeline''': | ||
* 2022-03-12 | * 2022-03-12: Data 4 Good kickoff event | ||
* 2022-03-16 project kickoff (first working session) | * 2022-03-16: Actual project kickoff (first working session) | ||
* 2022-06-12 | * 2022-06-12: Planned end of the project (Data 4 Good Demo Day) | ||
=== Resources / Contributing === | === Resources / Contributing === | ||
The weekly meeting for this project is | The weekly meeting for this project is Wednesdays at 20.00 (French time). We have a meeting room at https://meet.jit.si/DFG-OPENFOODFACTS | ||
You should also join | You should also join the Open Food Facts and Data 4 Good slacks. To be invited, use https://dataforgood.fr/join/ and https://slack.openfoodfacts.org | ||
'''Main | '''Main GitHub repositories''' | ||
* [https://github.com/openfoodfacts/robotoff/ robotoff]: | * [https://github.com/openfoodfacts/robotoff/ robotoff]: Robotoff, our program that handles predictions / insights | ||
* [https://github.com/openfoodfacts/robotoff-ann/ robotoff-ann] is a complement to robotoff that | * good to read : https://openfoodfacts.github.io/robotoff/introduction/architecture/ | ||
* | * [https://github.com/openfoodfacts/robotoff-ann/ robotoff-ann] is a complement to robotoff that focuses on logo detection and embedding (we use nearest neighbors to classify logos) | ||
* Last training of category model (using only title and ingredients as features): https://github.com/kulizhsy/off-category-classification | |||
* All trained models are published as "releases" on https://github.com/openfoodfacts/openfoodfacts-ai | * All trained models are published as "releases" on https://github.com/openfoodfacts/openfoodfacts-ai | ||
* [https://openfoodfacts.github.io/api-documentation/ | * [https://openfoodfacts.github.io/api-documentation/ Open Food Facts API documentation] | ||
=== Archives === | === Archives === | ||
* [https://docs.google.com/presentation/d/1fMN4di6AN1vz4sC3HYsA1PTya0mDGJYZ9htQYyqkLZQ/edit Data4Good launch presentation ( | * [https://docs.google.com/presentation/d/1fMN4di6AN1vz4sC3HYsA1PTya0mDGJYZ9htQYyqkLZQ/edit Data4Good launch presentation (Google Docs)] | ||
[[Category:Project]] | [[Category:Project]] | ||
[[Category:Robotoff]] | [[Category:Robotoff]] | ||
Revision as of 08:50, 16 March 2022
Summary
DataForGood is a French "association" that mobilizes tech to help citizen projects. We are part of season #10, this is our 4th season.
Main focus this year is on using machine learning to automatically categorize products so as to enable computing Nutri-Score and Éco-Score.
Description
Status: started
Expected outcomes: Have a new machine learning model, ready to deploy in production:
- That uses more features to predict category: OCR data, nutritional information, eventually images or any feature.
- Precision should be very high to be able to apply category automatically but the model should use its confidence to ask the user whenever needed.
Impact: massively increase the number of Eco-Scores and Nutri-Scores we are able to provide on the Open Food Facts platform.
Categorization is also important in many ways, for example to compare nutrition data with other products of same category (rank within a category).
Timeline:
- 2022-03-12: Data 4 Good kickoff event
- 2022-03-16: Actual project kickoff (first working session)
- 2022-06-12: Planned end of the project (Data 4 Good Demo Day)
Resources / Contributing
The weekly meeting for this project is Wednesdays at 20.00 (French time). We have a meeting room at https://meet.jit.si/DFG-OPENFOODFACTS
You should also join the Open Food Facts and Data 4 Good slacks. To be invited, use https://dataforgood.fr/join/ and https://slack.openfoodfacts.org
Main GitHub repositories
- robotoff: Robotoff, our program that handles predictions / insights
* good to read : https://openfoodfacts.github.io/robotoff/introduction/architecture/
- robotoff-ann is a complement to robotoff that focuses on logo detection and embedding (we use nearest neighbors to classify logos)
- Last training of category model (using only title and ingredients as features): https://github.com/kulizhsy/off-category-classification
- All trained models are published as "releases" on https://github.com/openfoodfacts/openfoodfacts-ai
- Open Food Facts API documentation