Questions tagged with Amazon Personalize

According to the documentation, with exploration weight as '0', there should be no exploration and the items returned should be based on relevance. With our testing we don't have the items returned completely based on relevance, it still returns popular items which the user has not interacted with at all. Our dataset has 50 users (with 100 interactions each) and also some 17 anonymous interactions (with no user_id and only session_id). Is this an issue with the dataset somehow? This is our user-item interaction schema example: ```{ "type": "record", "name": "Interactions", "namespace": "com.amazonaws.personalize.schema", "fields": [ { "name": "USER_ID", "type": "string" }, { "name": "ITEM_ID", "type": "string" }, { "name": "EVENT_TYPE", "type": "string" }, { "name": "DEVICE_TYPE", "type": "string" }, { "name": "LOCATION", "type": "string" }, { "name": "SESSION_ID", "type": [ "null", "string" ] }, { "name": "TIMESTAMP", "type": "long" } ] }lg...

For the User-Personalization recipe: According to this documentation: https://docs.aws.amazon.com/personalize/latest/dg/recording-events.html#recording-anonymous-user-events:~:text=You%20can%20record%20events%20for%20users%20before%20they%20create%20an%20account.%20This%20allows%20you%20to%20get%20recommendations%20for%20anonymous%20users.%20You%20can%20provide%20the%20sessionId%20as%20the%20userID%20in%20your%20GetRecommendations%20request. It seems like we should be able to query the Personalize engine for a `session_id` if it has interactions and does not have a `user_id`. However, when trying this and querying with a `session_id`, it still returns the high popularity items instead of relevant item recommendations. Is this incorrect, or will it only return relevant items after the user creates an account (e.g. has a `user_id` in the Users dataset)? Thanks in advancelg...

Does promotional filters apply only on the recommended items or does it always return a specific category of items in the recommendations? I am trying to apply a filter and % as 20 but it does't return any promotional recommendation?lg...

Hi, I am trying to figure out how to feed explicit negative feedback into AWSP. I have seen from previous questions and replies that is not possible to give negative feedback or weigh the events in a way that show that certain event is more important than another. Is my understanding correct? (I see that one of the suggestions is to use the negative feedback items for filtering only, whereas training should be on positive events only. As we already have a filtering functionality post-recommendation, so I am not interested in the filtering part) I can see the following options: - Include only interactions that received positive feedback in the interactions dataset - Include the whole interactions dataset, but train on positive interactions only (what will be the difference comparing it to the previous case?) - Artificially create multiple entries for an interaction record (e.g. 5 for positive feedback, 1 for negative, this way showing preference for the item) - Use impressions (with the negative feedback items shown as displayed, but not selected, to indicate less relevance) Which way would work better? Has anyone tried any of these? Any pitfalls with any of those options? Thank you very much for your help!lg...

For example, if I have the following User schema: ``` { "name": "USER_ID", "type": "string" }, { "name": "AGE", "type": [ "null", "int" ], "categorical": true }, { "name": "GENDER", "type": [ "null", "string" ], }, { "name": "USER_DESCRIPTION", "type": "string" } ``` Will the "GENDER" field be used for training or will only the "AGE" be used for training? Also, what are the differences between contextual & categorial metadata? Is this applicable for Users, Item, and User-Item interactions? I am new to Amazon Personalize, so I apologize if this is trivial.lg...

Hello there - testing our Amazon Personalize for a business use-case where we want to suggest other Users with similar Items/Events. Based on research it seems like a user-user recommendation system isn't possible with Amazon Personalize, but wanted to double check to make sure I'm not overlooking a creative solution for this. Thank you!lg...

For **user-personalization recipe**, I have the following metadata: * **Users:** User_id, age, gender, * **Interactions:** User_id, Item_id, event_type, * **Items:** Item_id, Genre and Item_title. We want to get the recommendations filtered based on the age. Currently in the create filter step, we aren't able to add Users.age property in our filter for the where clause. Any suggestions on how we can filter on different users metadata for the user personalization recipe?lg...

Getting this error when trying to add filter on other metadata for interactions dataset. "**Could not create filter: Interaction dataset supports condition on only event_type field.**" Can anyone confirm if we can add the filter on only the event_type for interactions?lg...

I am having trouble with the deletion of Interactions Dataset. I was trying to delete the existing Dataset in my Dataset Group, in order to create a new one, with the boto3 delete_dataset() function (from a Sagemaker Notebook) when I recieved the following error: "ResourceInUseException: An error occurred (ResourceInUseException) when calling the DeleteDataset operation: Please delete all filters before deleting the dataset". I have no filters in the datasets, that is why I was confused. I tried also to delete the same dataset using the CLI, but recieved the same error message. I stoped the ACTIVE recommender in case there was a problem with that but the same error occurred. Can you help me with a solution for this error? Thank you in advance.lg...

Is there a way that I can provide a callback function that executes when a solution version changes state to 'active' ? Currently, my system successfully creates a new solution version, but I want to automatically do additional processing elsewhere when the solution version is active. I have looked over the following documentation articles: * https://docs.aws.amazon.com/personalize/latest/dg/creating-a-solution-version.html * https://docs.aws.amazon.com/AWSJavaScriptSDK/latest/AWS/Personalize.html#createSolutionVersion-property The only way that I can currently think about solving this is using setTimeout (for let's say 15 minutes), calling DescribeSolutionVersion, if status is active -> doAdditionalProcessing() else -> setTimeout for another 15 minutes. Is there a better way of going about this? I am new to AWS I apologize if this is triviallg...

I am executing filter to Perzonalize, where the filterValues returns a message that the maximum length is 1000 and my filter has a length of 6664. what can I do about it so that it takes the full length of my filter? Documentation: [https://docs.aws.amazon.com/personalize/latest/dg/filter-real-time.html]() Error message: ``` at 'filterValues' failed to satisfy constraint: Map value must satisfy constraint: [Member must have length less than or equal to 1000, Member must have length greater than or equal to 0]\n at Object.extractError (/var/runtime/node_modules/aws-sdk/lib/protocol/json.js:52:27) ```lg...

What is the model family/type behind Amazon Personalize's User-Personalization recipe?lg...