docs: Document storage implications for training data

Contributes to: #172

Signed-off-by: Dale Lane <[email protected]>

public/components/help/help.html

@@ -379,6 +379,22 @@ <h4 class="panel-title">
                 </div>
             </div>
         </div>
+        <div class="panel panel-default">
+            <div class="panel-heading">
+                <h4 class="panel-title">
+                    <a data-toggle="collapse" href="#helpTrainingData" target="_self" translate="HELP.ACCOUNTISSUES.Q5"></a>
+                </h4>
+            </div>
+            <div id="helpTrainingData" class="panel-collapse collapse">
+                <div class="panel-body">
+                    <p translate="HELP.ACCOUNTISSUES.Q5-A-1"></p>
+                    <p translate="HELP.ACCOUNTISSUES.Q5-A-2"></p>
+                    <p translate="HELP.ACCOUNTISSUES.Q5-A-3"></p>
+                    <p translate="HELP.ACCOUNTISSUES.Q5-A-4"></p>
+                    <p translate="HELP.ACCOUNTISSUES.Q5-A-5"></p>
+                </div>
+            </div>
+        </div>
     </div>
 
 </div>

public/languages/de.json

@@ -1007,7 +1007,14 @@
             "Q4-A-6": "Allgemeine Information, die in allen Accounts gespeichert wird",
             "Q4-A-7": "Für das Benutzermanagement bei Machine Learnign for Kids wird der Service <a href='https://auth0.com/'>Auth0</a> genutzt. Auth0 speichert die IP Adresse und den Browser, den du benutzt.",
             "Q4-A-8": "Um Fehler im Webbrowser festzustellen, nutzt Machine Learning for Kids den Service <a href='https://sentry.io/'>Sentry</a>. Wenn ein Fehler auftritt, werde der Benutzername, die IP Adresse, die Browserart und eine technische Beschreibung des Fehlers, gesammelt.",
-            "Q4-A-9": "Wir nutzen <a href='https://www.google.com/analytics/'>Google Analytics</a>, um einen Überblick darüber zu erhalten, wie viele Nutzer unsere Plattform - Machine Learning for Kids - jeden Tag besuchen. Getrackte geographische Informationen sowie Informationen über den genutzten Browser werden nur in anonymisierter Form angezeigt."
+            "Q4-A-9": "Wir nutzen <a href='https://www.google.com/analytics/'>Google Analytics</a>, um einen Überblick darüber zu erhalten, wie viele Nutzer unsere Plattform - Machine Learning for Kids - jeden Tag besuchen. Getrackte geographische Informationen sowie Informationen über den genutzten Browser werden nur in anonymisierter Form angezeigt.",
+
+            "Q5": "What happens to training data created by students?",
+            "Q5-A-1": "A step in all machine learning projects that students do using this site is to create or collect training examples. For text projects, this can mean typing in short text passages. For image projects, this can means drawing pictures onto an on-screen canvas or taking photos using the webcam. For numbers projects, this means typing in sets of numbers",
+            "Q5-A-2": "All of these training examples are stored in a <a href='https://www.ibm.com/cloud/compose/mysql'>MySQL database hosted on IBM Cloud</a>. As the database administrator, I have access to all of these training examples for all student projects, however I do not use it for any other purposes other than enabling students to access and work on their own projects. I do not publish the training data (either publicly or to individuals or third parties) or otherwise share it, with the two exceptions of submitting it to IBM Watson APIs described below.",
+            "Q5-A-3": "Machine learning models for text projects are created using the <a href='https://www.ibm.com/cloud/watson-assistant/'>IBM Watson Assistant</a> cloud service, which means when students train a model, a copy of the training examples that they have entered will be <a href='https://cloud.ibm.com/apidocs/assistant#create-workspace'>submitted to IBM Watson</a>. IBM Watson does not use these training examples for any other purposes. This is because Machine Learning for Kids <a href='https://cloud.ibm.com/apidocs/assistant#data-collection'>requests an opt-out</a> for student training examples being used by IBM to improve the service, as <a href='https://cloud.ibm.com/docs/services/watson?topic=watson-gs-logging-overview'>described in the Watson documentation</a>.",
+            "Q5-A-4": "Machine learning models for images projects are created using the <a href='https://www.ibm.com/watson/services/visual-recognition/'>IBM Visual Recognition</a> cloud service, which means when students train a model, a copy of their training images will be <a href='https://cloud.ibm.com/apidocs/visual-recognition#create-a-classifier'>uploaded to IBM Watson</a>. IBM Watson does not use these training examples for any other purposes. This is because Machine Learning for Kids <a href='https://cloud.ibm.com/apidocs/visual-recognition#data-collection'>requests an opt-out</a> for student training examples being used by IBM to improve the service, as <a href='https://cloud.ibm.com/docs/services/watson?topic=watson-gs-logging-overview'>described in the Watson documentation</a>.",
+            "Q5-A-5": "IBM Watson provides some <a href='https://cloud.ibm.com/docs/services/assistant?topic=assistant-information-security#information-security-gdpr-wa'>general guidance for sharing training data with their services</a>, which also apply to training data created and collected in Machine Learning for Kids. Specifically, students should be advised not to include personally identifiable information in the examples that they collect."
         }
     }
 }

public/languages/en.json

@@ -1007,7 +1007,14 @@
             "Q4-A-6": "General information stored about all accounts",
             "Q4-A-7": "User management for Machine Learning for Kids is implemented using the third party service, <a href='https://auth0.com/'>Auth0</a>. They store the IP address that you last logged into Machine Learning for Kids from, and the type of browser you used. I've never found a reason to use that, but it is stored if I did want to go and look for it.",
             "Q4-A-8": "Errors that happen in the web browser are captured using the third party service, <a href='https://sentry.io/'>Sentry</a>. If something goes wrong, it will capture information about the error, including your username, IP address, type of browser you were using, and a technical description of what went wrong.",
-            "Q4-A-9": "I use <a href='https://www.google.com/analytics/'>Google Analytics</a> so that I know how many users visit Machine Learning for Kids each day. Although it captures information such as geographic location and browser type, this is only ever displayed to me in an anonymised aggregate way."
+            "Q4-A-9": "I use <a href='https://www.google.com/analytics/'>Google Analytics</a> so that I know how many users visit Machine Learning for Kids each day. Although it captures information such as geographic location and browser type, this is only ever displayed to me in an anonymised aggregate way.",
+
+            "Q5": "What happens to training data created by students?",
+            "Q5-A-1": "A step in all machine learning projects that students do using this site is to create or collect training examples. For text projects, this can mean typing in short text passages. For image projects, this can means drawing pictures onto an on-screen canvas or taking photos using the webcam. For numbers projects, this means typing in sets of numbers",
+            "Q5-A-2": "All of these training examples are stored in a <a href='https://www.ibm.com/cloud/compose/mysql'>MySQL database hosted on IBM Cloud</a>. As the database administrator, I have access to all of these training examples for all student projects, however I do not use it for any other purposes other than enabling students to access and work on their own projects. I do not publish the training data (either publicly or to individuals or third parties) or otherwise share it, with the two exceptions of submitting it to IBM Watson APIs described below.",
+            "Q5-A-3": "Machine learning models for text projects are created using the <a href='https://www.ibm.com/cloud/watson-assistant/'>IBM Watson Assistant</a> cloud service, which means when students train a model, a copy of the training examples that they have entered will be <a href='https://cloud.ibm.com/apidocs/assistant#create-workspace'>submitted to IBM Watson</a>. IBM Watson does not use these training examples for any other purposes. This is because Machine Learning for Kids <a href='https://cloud.ibm.com/apidocs/assistant#data-collection'>requests an opt-out</a> for student training examples being used by IBM to improve the service, as <a href='https://cloud.ibm.com/docs/services/watson?topic=watson-gs-logging-overview'>described in the Watson documentation</a>.",
+            "Q5-A-4": "Machine learning models for images projects are created using the <a href='https://www.ibm.com/watson/services/visual-recognition/'>IBM Visual Recognition</a> cloud service, which means when students train a model, a copy of their training images will be <a href='https://cloud.ibm.com/apidocs/visual-recognition#create-a-classifier'>uploaded to IBM Watson</a>. IBM Watson does not use these training examples for any other purposes. This is because Machine Learning for Kids <a href='https://cloud.ibm.com/apidocs/visual-recognition#data-collection'>requests an opt-out</a> for student training examples being used by IBM to improve the service, as <a href='https://cloud.ibm.com/docs/services/watson?topic=watson-gs-logging-overview'>described in the Watson documentation</a>.",
+            "Q5-A-5": "IBM Watson provides some <a href='https://cloud.ibm.com/docs/services/assistant?topic=assistant-information-security#information-security-gdpr-wa'>general guidance for sharing training data with their services</a>, which also apply to training data created and collected in Machine Learning for Kids. Specifically, students should be advised not to include personally identifiable information in the examples that they collect."
         }
     }
 }

public/languages/es.json

@@ -1007,7 +1007,14 @@
 			"Q4-A-6": "Información general sobre todas las cuentas. ",
 			"Q4-A-7": "La gestión de usuarios de  Machine Learning for Kids se lleva a cabo mediante un Servicio de terceros, <a href='https://auth0.com/'>Auth0</a>. Almacenan la dirección IP del ultimo inicio de session y el tipo de navegador usado. Nunca he encontrado una razón para usarlo, pero se guarda por si resulta necesario..",
 			"Q4-A-8": "Los errors del navegador se recogen mediante un servicio de tercerosa <a href='https://sentry.io/'>Sentry</a>. Si algo falla, recogerá la información del error, incluyendo nombre de usurio, dirección IP, tipo de navegador y una descripción técnica de lo que ha fallado.",
-			"Q4-A-9": "Uso <a href='https://www.google.com/analytics/'>Google Analytics</a> para saber cuántos usuarios visitan visit Machine Learning for Kids cada día. Aunque también recoge información geográfica o tipo de navegador, esto solo puedo verlo de forma agregada y anonimizada."
+			"Q4-A-9": "Uso <a href='https://www.google.com/analytics/'>Google Analytics</a> para saber cuántos usuarios visitan visit Machine Learning for Kids cada día. Aunque también recoge información geográfica o tipo de navegador, esto solo puedo verlo de forma agregada y anonimizada.",
+
+            "Q5": "What happens to training data created by students?",
+            "Q5-A-1": "A step in all machine learning projects that students do using this site is to create or collect training examples. For text projects, this can mean typing in short text passages. For image projects, this can means drawing pictures onto an on-screen canvas or taking photos using the webcam. For numbers projects, this means typing in sets of numbers",
+            "Q5-A-2": "All of these training examples are stored in a <a href='https://www.ibm.com/cloud/compose/mysql'>MySQL database hosted on IBM Cloud</a>. As the database administrator, I have access to all of these training examples for all student projects, however I do not use it for any other purposes other than enabling students to access and work on their own projects. I do not publish the training data (either publicly or to individuals or third parties) or otherwise share it, with the two exceptions of submitting it to IBM Watson APIs described below.",
+            "Q5-A-3": "Machine learning models for text projects are created using the <a href='https://www.ibm.com/cloud/watson-assistant/'>IBM Watson Assistant</a> cloud service, which means when students train a model, a copy of the training examples that they have entered will be <a href='https://cloud.ibm.com/apidocs/assistant#create-workspace'>submitted to IBM Watson</a>. IBM Watson does not use these training examples for any other purposes. This is because Machine Learning for Kids <a href='https://cloud.ibm.com/apidocs/assistant#data-collection'>requests an opt-out</a> for student training examples being used by IBM to improve the service, as <a href='https://cloud.ibm.com/docs/services/watson?topic=watson-gs-logging-overview'>described in the Watson documentation</a>.",
+            "Q5-A-4": "Machine learning models for images projects are created using the <a href='https://www.ibm.com/watson/services/visual-recognition/'>IBM Visual Recognition</a> cloud service, which means when students train a model, a copy of their training images will be <a href='https://cloud.ibm.com/apidocs/visual-recognition#create-a-classifier'>uploaded to IBM Watson</a>. IBM Watson does not use these training examples for any other purposes. This is because Machine Learning for Kids <a href='https://cloud.ibm.com/apidocs/visual-recognition#data-collection'>requests an opt-out</a> for student training examples being used by IBM to improve the service, as <a href='https://cloud.ibm.com/docs/services/watson?topic=watson-gs-logging-overview'>described in the Watson documentation</a>.",
+            "Q5-A-5": "IBM Watson provides some <a href='https://cloud.ibm.com/docs/services/assistant?topic=assistant-information-security#information-security-gdpr-wa'>general guidance for sharing training data with their services</a>, which also apply to training data created and collected in Machine Learning for Kids. Specifically, students should be advised not to include personally identifiable information in the examples that they collect."
 		}
 	}
 }

public/languages/fr.json

@@ -1007,7 +1007,14 @@
             "Q4-A-6": "Informations générales stockées sur tous les comptes",
             "Q4-A-7": "La gestion des utilisateurs pour Machine Learning for Kids est implémentée en utilisant le service tiers, <a href='https://auth0.com/'>Auth0</a>. Ils enregistrent l'adresse IP à partir de laquelle vous vous êtes connecté pour la dernière fois à Machine Learning for Kids et le type de navigateur que vous avez utilisé. Je n'ai jamais trouvé de raison de l'utiliser, mais il est stocké si je veux aller le chercher.",
             "Q4-A-8": "Les erreurs qui se produisent dans le navigateur Web sont capturées en utilisant le service tiers, <a href='https://sentry.io/'>Sentry</a>. Si quelque chose ne va pas, il saisira des informations sur l'erreur, y compris votre nom d'utilisateur, votre adresse IP, le type de navigateur que vous utilisiez et une description technique de ce qui a dysfonctionné.",
-            "Q4-A-9": "J'utilise <a href='https://www.google.com/analytics/'>Google Analytics</a> pour savoir combien d'utilisateurs visitent Machine Learning for Kids chaque jour. Bien qu'il capture des informations telles que l'emplacement géographique et le type de navigateur, il ne me les affiche uniquement que sous forme d'agrégat anonyme."
+            "Q4-A-9": "J'utilise <a href='https://www.google.com/analytics/'>Google Analytics</a> pour savoir combien d'utilisateurs visitent Machine Learning for Kids chaque jour. Bien qu'il capture des informations telles que l'emplacement géographique et le type de navigateur, il ne me les affiche uniquement que sous forme d'agrégat anonyme.",
+
+            "Q5": "What happens to training data created by students?",
+            "Q5-A-1": "A step in all machine learning projects that students do using this site is to create or collect training examples. For text projects, this can mean typing in short text passages. For image projects, this can means drawing pictures onto an on-screen canvas or taking photos using the webcam. For numbers projects, this means typing in sets of numbers",
+            "Q5-A-2": "All of these training examples are stored in a <a href='https://www.ibm.com/cloud/compose/mysql'>MySQL database hosted on IBM Cloud</a>. As the database administrator, I have access to all of these training examples for all student projects, however I do not use it for any other purposes other than enabling students to access and work on their own projects. I do not publish the training data (either publicly or to individuals or third parties) or otherwise share it, with the two exceptions of submitting it to IBM Watson APIs described below.",
+            "Q5-A-3": "Machine learning models for text projects are created using the <a href='https://www.ibm.com/cloud/watson-assistant/'>IBM Watson Assistant</a> cloud service, which means when students train a model, a copy of the training examples that they have entered will be <a href='https://cloud.ibm.com/apidocs/assistant#create-workspace'>submitted to IBM Watson</a>. IBM Watson does not use these training examples for any other purposes. This is because Machine Learning for Kids <a href='https://cloud.ibm.com/apidocs/assistant#data-collection'>requests an opt-out</a> for student training examples being used by IBM to improve the service, as <a href='https://cloud.ibm.com/docs/services/watson?topic=watson-gs-logging-overview'>described in the Watson documentation</a>.",
+            "Q5-A-4": "Machine learning models for images projects are created using the <a href='https://www.ibm.com/watson/services/visual-recognition/'>IBM Visual Recognition</a> cloud service, which means when students train a model, a copy of their training images will be <a href='https://cloud.ibm.com/apidocs/visual-recognition#create-a-classifier'>uploaded to IBM Watson</a>. IBM Watson does not use these training examples for any other purposes. This is because Machine Learning for Kids <a href='https://cloud.ibm.com/apidocs/visual-recognition#data-collection'>requests an opt-out</a> for student training examples being used by IBM to improve the service, as <a href='https://cloud.ibm.com/docs/services/watson?topic=watson-gs-logging-overview'>described in the Watson documentation</a>.",
+            "Q5-A-5": "IBM Watson provides some <a href='https://cloud.ibm.com/docs/services/assistant?topic=assistant-information-security#information-security-gdpr-wa'>general guidance for sharing training data with their services</a>, which also apply to training data created and collected in Machine Learning for Kids. Specifically, students should be advised not to include personally identifiable information in the examples that they collect."
         }
     }
 }