Difference between revisions of "Artificial Intelligence/Open Products Facts"

From Open Food Facts wiki
Jump to navigation Jump to search
(Created page with "=== ✅ Tasks === * Plug OPF Redis to Robotoff * Retrain category model (use the name-based simplistic model temporarily ?) * Adapt Hunger Games logo dashboard to be generic (...")
 
 
(4 intermediate revisions by the same user not shown)
Line 1: Line 1:
−
=== ✅ Tasks ===
+
=== ✅ Priority tasks ===
−
* Plug OPF Redis to Robotoff
+
* Plug the Open Products Facts Redis to Robotoff
−
* Retrain category model (use the name-based simplistic model temporarily ?)
+
* Retrain our category prediction model on Open Products Facts (use the name-based simplistic model temporarily ?)
−
* Adapt Hunger Games logo dashboard to be generic (based on existing labels logos)
+
* Adapt our Hunger Games logo dashboard to be generic (based on existing labels logos)
−
* Ensure Robotoff can use OPF specific categories
+
* Ensure Robotoff can use Open Products Facts specific categories
−
 
+
* Classifier for product type (food, petfood, cosmetic, other)
  
 
=== GitHub ===
 
=== GitHub ===
 
* https://github.com/openfoodfacts/robotoff/labels/%F0%9F%93%B8%20Open%20Products%20Facts
 
* https://github.com/openfoodfacts/robotoff/labels/%F0%9F%93%B8%20Open%20Products%20Facts
 
* https://github.com/openfoodfacts/openfoodfacts-ai/labels/%F0%9F%93%B8%20Open%20Products%20Facts
 
* https://github.com/openfoodfacts/openfoodfacts-ai/labels/%F0%9F%93%B8%20Open%20Products%20Facts
 +
[[Category:Open Products Facts]]
 +
[[Category:Artificial Intelligence]]
 +
 +
 +
=== Using GenAI to generate properties to describe categories ===
 +
Considerations:
 +
* the prompt should be edited to output in folksonomy compliant format, include units, descriptions, create a wiki page for the property, add constraints
 +
 +
<pre>
 +
Can you list as bullet points the relevant data fields to describe "Molecular diagnostics analyzer" for a product datasheet.
 +
- Product Name: Molecular Diagnostics Analyzer
 +
- Application: [specify the type of sample analysis, such as clinical or research]
 +
- Technology: [indicate the molecular diagnostic technology used, such as PCR, NGS, etc.]
 +
- Sample type compatibility: [list the types of samples that can be analyzed, such as blood, saliva, etc.]
 +
- Detection methods: [indicate the types of targets the analyzer can detect, such as DNA, RNA, pathogens, etc.]
 +
- Throughput: [indicate the number of samples that can be processed per unit of time, such as per hour or day]
 +
- Run time: [specify the total time required for a sample analysis, from sample preparation to result generation]
 +
- Sensitivity: [indicate the minimum concentration of the target that can be detected]
 +
- Specificity: [indicate the degree of accuracy in detecting the target of interest, as well as avoiding false positive results]
 +
- Data management: [indicate the type of data storage and analysis capabilities, such as cloud storage, real-time data analysis, etc.]
 +
- User interface: [describe the type of interface for sample analysis, such as touchscreen or computer-based]
 +
- Power source: [indicate the type of power required, such as AC power or battery-operated]
 +
- Dimensions: [provide the physical dimensions of the analyzer, such as height, width, and depth]
 +
- Weight: [indicate the weight of the analyzer for ease of transport and handling]
 +
- Brand: [name of the brand producing the product]
 +
- SKU or product code: [unique identifier for the product]
 +
- MSRP: [manufacturer's suggested retail price].
 +
</pre>
 +
 +
=== Using GenAI to generate service provider descriptions ===
 +
<pre>
 +
Generate a config file/ontology to describe all the circular services that help through the life cycle of any product. Topics of services might include Spare parts, donation, reparation,  Resale,  environmental scoring methodology algorithm (please include a way to call the algorithm from an external file), purchase decision, P2P lending… You should take into account the fact that all those services are not available in all countries, that all services are not available for all product categories. Each of those services might have a website or an app or an API, or all three of them (do include example of urls).  You can take a few concrete examples to make it more Concrete, using textile and electronics as sample categories.  Please include the name of those circular services.
 +
 +
 +
 +
 +
services:
 +
  - name: Spare parts
 +
      available_countries: [USA, Germany, France]
 +
      available_categories: [Textile, Electronics]
 +
      website: https://www.spareparts.com
 +
 +
 +
  - name: Donation
 +
      available_countries: [USA, Canada, UK]
 +
      available_categories: [Textile, Electronics]
 +
      website: https://www.donationhub.com
 +
 +
 +
  - name: Reparation
 +
      available_countries: [USA, Spain, Australia]
 +
      available_categories: [Electronics]
 +
      website: https://www.repairnow.com
 +
 +
 +
  - name: Resale
 +
      available_countries: [USA, UK, Japan]
 +
      available_categories: [Textile, Electronics]
 +
      website: https://www.resalemarketplace.com
 +
 +
 +
  - name: Environmental scoring methodology algorithm
 +
      available_countries: [USA, Germany, Canada]
 +
      available_categories: [Textile, Electronics]
 +
      algorithm_file: scoring_algorithm.py
 +
 +
 +
  - name: Purchase decision
 +
      available_countries: [USA, France, Australia]
 +
      available_categories: [Textile, Electronics]
 +
      website: https://www.purchaseadvisor.com
 +
 +
 +
  - name: P2P lending
 +
      available_countries: [USA, UK, Japan]
 +
      available_categories: [Textile, Electronics]
 +
      website: https://www.p2plendingplatform.com
 +
 +
</pre>
 +
 +
<pre>
 +
Generate a config file/ontology to describe all the circular services that help through the life cycle of any product. Topics of services might include Spare parts, donation, reparation,  Resale,  purchase decision, P2P lending… You should take into account the fact that all those services are not available in all countries, that all services are not available for all product categories. Each of those services might have a website or an app or an API, or all three of them (do include example of urls).  You can take a few concrete examples to make it more Concrete, using textile and electronics as sample categories.
 +
 +
 +
 +
 +
categories:
 +
  - name: Textile
 +
      services:
 +
          - name: Spare Parts
 +
              available_countries:
 +
                  - USA
 +
                  - Germany
 +
              websites:
 +
                  - url: https://www.textilespareparts.com
 +
                      type: website
 +
                  - url: https://api.textilespareparts.com
 +
                      type: api
 +
          - name: Donation
 +
              available_countries:
 +
                  - USA
 +
                  - UK
 +
              websites:
 +
                  - url: https://www.textiledonation.org
 +
                      type: website
 +
                  - url: https://api.textiledonation.org
 +
                      type: api
 +
          - name: Reparation
 +
              available_countries:
 +
                  - USA
 +
                  - France
 +
              websites:
 +
                  - url: https://www.textilereparation.com
 +
                      type: website
 +
                  - url: https://api.textilereparation.com
 +
                      type: api
 +
 +
 +
  - name: Electronics
 +
      services:
 +
          - name: Spare Parts
 +
              available_countries:
 +
                  - USA
 +
                  - Japan
 +
              websites:
 +
                  - url: https://www.electronicsspareparts.com
 +
                      type: website
 +
                  - url: https://api.electronicsspareparts.com
 +
                      type: api
 +
          - name: Resale
 +
              available_countries:
 +
                  - USA
 +
                  - Canada
 +
              websites:
 +
                  - url: https://www.electronicsresale.com
 +
                      type: website
 +
                  - url: https://api.electronicsresale.com
 +
                      type: api
 +
          - name: Purchase Decision
 +
              available_countries:
 +
                  - USA
 +
                  - UK
 +
              websites:
 +
                  - url: https://www.electronicsdecision.com
 +
                      type: website
 +
                  - url: https://api.electronicsdecision.com
 +
                      type: api
 +
 +
</pre>

Latest revision as of 08:42, 6 September 2024

✅ Priority tasks

  • Plug the Open Products Facts Redis to Robotoff
  • Retrain our category prediction model on Open Products Facts (use the name-based simplistic model temporarily ?)
  • Adapt our Hunger Games logo dashboard to be generic (based on existing labels logos)
  • Ensure Robotoff can use Open Products Facts specific categories
  • Classifier for product type (food, petfood, cosmetic, other)

GitHub


Using GenAI to generate properties to describe categories

Considerations:

  • the prompt should be edited to output in folksonomy compliant format, include units, descriptions, create a wiki page for the property, add constraints
Can you list as bullet points the relevant data fields to describe "Molecular diagnostics analyzer" for a product datasheet.
- Product Name: Molecular Diagnostics Analyzer
- Application: [specify the type of sample analysis, such as clinical or research]
- Technology: [indicate the molecular diagnostic technology used, such as PCR, NGS, etc.]
- Sample type compatibility: [list the types of samples that can be analyzed, such as blood, saliva, etc.]
- Detection methods: [indicate the types of targets the analyzer can detect, such as DNA, RNA, pathogens, etc.]
- Throughput: [indicate the number of samples that can be processed per unit of time, such as per hour or day]
- Run time: [specify the total time required for a sample analysis, from sample preparation to result generation]
- Sensitivity: [indicate the minimum concentration of the target that can be detected]
- Specificity: [indicate the degree of accuracy in detecting the target of interest, as well as avoiding false positive results]
- Data management: [indicate the type of data storage and analysis capabilities, such as cloud storage, real-time data analysis, etc.]
- User interface: [describe the type of interface for sample analysis, such as touchscreen or computer-based]
- Power source: [indicate the type of power required, such as AC power or battery-operated]
- Dimensions: [provide the physical dimensions of the analyzer, such as height, width, and depth]
- Weight: [indicate the weight of the analyzer for ease of transport and handling]
- Brand: [name of the brand producing the product]
- SKU or product code: [unique identifier for the product]
- MSRP: [manufacturer's suggested retail price].

Using GenAI to generate service provider descriptions

Generate a config file/ontology to describe all the circular services that help through the life cycle of any product. Topics of services might include Spare parts, donation, reparation,  Resale,  environmental scoring methodology algorithm (please include a way to call the algorithm from an external file), purchase decision, P2P lending… You should take into account the fact that all those services are not available in all countries, that all services are not available for all product categories. Each of those services might have a website or an app or an API, or all three of them (do include example of urls).  You can take a few concrete examples to make it more Concrete, using textile and electronics as sample categories.  Please include the name of those circular services.




services:
   - name: Spare parts
       available_countries: [USA, Germany, France]
       available_categories: [Textile, Electronics]
       website: https://www.spareparts.com


   - name: Donation
       available_countries: [USA, Canada, UK]
       available_categories: [Textile, Electronics]
       website: https://www.donationhub.com


   - name: Reparation
       available_countries: [USA, Spain, Australia]
       available_categories: [Electronics]
       website: https://www.repairnow.com


   - name: Resale
       available_countries: [USA, UK, Japan]
       available_categories: [Textile, Electronics]
       website: https://www.resalemarketplace.com


   - name: Environmental scoring methodology algorithm
       available_countries: [USA, Germany, Canada]
       available_categories: [Textile, Electronics]
       algorithm_file: scoring_algorithm.py


   - name: Purchase decision
       available_countries: [USA, France, Australia]
       available_categories: [Textile, Electronics]
       website: https://www.purchaseadvisor.com


   - name: P2P lending
       available_countries: [USA, UK, Japan]
       available_categories: [Textile, Electronics]
       website: https://www.p2plendingplatform.com


Generate a config file/ontology to describe all the circular services that help through the life cycle of any product. Topics of services might include Spare parts, donation, reparation,  Resale,  purchase decision, P2P lending… You should take into account the fact that all those services are not available in all countries, that all services are not available for all product categories. Each of those services might have a website or an app or an API, or all three of them (do include example of urls).  You can take a few concrete examples to make it more Concrete, using textile and electronics as sample categories. 




categories:
   - name: Textile
       services:
           - name: Spare Parts
               available_countries:
                   - USA
                   - Germany
               websites:
                   - url: https://www.textilespareparts.com
                       type: website
                   - url: https://api.textilespareparts.com
                       type: api
           - name: Donation
               available_countries:
                   - USA
                   - UK
               websites:
                   - url: https://www.textiledonation.org
                       type: website
                   - url: https://api.textiledonation.org
                       type: api
           - name: Reparation
               available_countries:
                   - USA
                   - France
               websites:
                   - url: https://www.textilereparation.com
                       type: website
                   - url: https://api.textilereparation.com
                       type: api


   - name: Electronics
       services:
           - name: Spare Parts
               available_countries:
                   - USA
                   - Japan
               websites:
                   - url: https://www.electronicsspareparts.com
                       type: website
                   - url: https://api.electronicsspareparts.com
                       type: api
           - name: Resale
               available_countries:
                   - USA
                   - Canada
               websites:
                   - url: https://www.electronicsresale.com
                       type: website
                   - url: https://api.electronicsresale.com
                       type: api
           - name: Purchase Decision
               available_countries:
                   - USA
                   - UK
               websites:
                   - url: https://www.electronicsdecision.com
                       type: website
                   - url: https://api.electronicsdecision.com
                       type: api
Retrieved from "https://wiki.openfoodfacts.org/index.php?title=Artificial_Intelligence/Open_Products_Facts&oldid=23949"