Release Summary - Apr 18, 2024 (24.08)
The following key features and improvements, along with bug fixes, have been released in Algonomy CXP products in the release version 24.08 during Apr 05, 2024 - Apr 18, 2024.
Enterprise Dashboard
Auto-complete Functionality for User Segments and User Attributes
Merchandisers can now benefit from auto-complete suggestions when specifying user segments and attributes in settings for recommendation restrictions, boosting, manual recommendations, strategy rules, and segment builders. Whether adjusting context in strategy rules or defining shopper profiles, the system will provide a dynamic list of suggestions that match the user’s input, displaying both names and values where applicable.
Jira: ENG-26878
Configurable Strategies - show only whitelisted attributes for AttributeTopSellers
In Configurable Strategies, the Attribute Top Sellers and Attribute Top Views models now require attributes to be whitelisted before they can be used. Previously, users could select any attribute when creating a Configurable Strategy, which led to instances where unwhitelisted attributes were chosen, resulting in no recommendations being generated.
To address this, we have updated the UI so that only whitelisted attributes are available in the Product Attribute dropdown for the AttributeTopSellersModel. This ensures all attributes used in strategies are pre-approved and effective, enhancing the reliability and performance of recommendations generated through these models.
Jira: ENG-27868
Enhanced Currency Selection Persistence Across Content Report Tabs
To streamline the user experience for merchandisers navigating through various content report tabs, we have enhanced our platform to persist the "All Currencies" selection across different tabs. This improvement means that once you select "All Currencies" or any specific currency, the platform will remember your choice as you switch between different report tabs, provided the metric is available for those tabs. This feature aims to save time and improve workflow efficiency by eliminating the need to repeatedly select the currency option. This update is particularly useful for users working with reports in multiple currencies, ensuring a consistent and seamless analytical experience.
Jira: ENG-27866
Enhancements to Primary Category Configuration Page
We have updated the Primary Category Configuration page to be customer-facing and have resolved an issue that was causing errors during the publishing of selected categories. This improvement ensures a smoother and more intuitive user experience when managing primary category settings.
Jira: ENG-27911
Data Engineering
FIND: Add Find API fields to Visits
We have expanded the data collection in our Visits logs by integrating specific fields from the Find API. This improvement is designed to enhance reporting capabilities on search behaviors, including facets and zero search terms encounters by users.
The update to Visits now includes:
-
Date and time for each search event, capturing when a search term was entered, or a filter was applied.
-
Session ID and User ID to link search actions to specific visits.
-
Search term used by the visitor.
-
Filter applied during the search to refine results.
-
Number of results returned for the search term or after applying the filter.
Jira: ENG-27614
Other Feature Enhancements
The following feature enhancements and upgrades have been made in the release version 24.08 during Apr 05, 2024 - Apr 18, 2024.
|
Jira # |
Module/Title |
Summary |
General Availability |
|---|---|---|---|
|
|
Find: Dashboard backend changes to save rules to playservice |
We have developed new backend changes for the dashboard that facilitate the creation, modification, deletion, and retrieval of rules via the UI. These changes include a set of CRUD (Create, Read, Update, Delete) APIs that serve as a wrapper over the Play Service APIs, ensuring seamless interactions and data consistency. |
18-Apr-2024 |
|
|
Science: Catalog Management Enhancement - Site-Specific Schema Implementation |
We have optimized our catalog management system by transitioning to a site-specific schema to address scalability and performance challenges. This update involves migrating existing data to new site-specific tables and ensuring that all catalog enrichment processes, such as those run by data science and Apache Airflow jobs, are updated to interact with these new tables seamlessly. |
18-Apr-2024 |
|
|
Recommend: Make Recs.algorecs.com at par with recs.richrelevance.com for vanity URL config |
We have updated the vanity URL configurations to ensure consistent behavior across recs.richrelevance.com and algorecs.com domains. Previously, the platform-specific settings caused integration requests on algorecs.com to default to production URL responses, which could lead to configuration errors. With the new enhancement, enabling the 'response domain based on request domain' setting allows the RRserver to correctly respond with domain-specific URLs—integration requests return integration.algorecs.com and production requests return recs.algorecs.com. |
18-Apr-2024 |
|
|
Streaming Catalog: Custom Batch Size Configuration for Enrichment Calculations |
We've updated the enrichment calculation process to include a customizable batch size property. Previously, a universal batch size of 100 was used, which was reduced to 10 for some sites to avoid exceeding Kafka's message size limits. This change, while necessary, increased network traffic for other sites. With the new update, sites can now set their own batch sizes via new REST endpoints, optimizing performance without impacting network efficiency. The default batch size will remain at 100, allowing flexibility and improved data handling across different sites. |
18-Apr-2024 |
|
|
Use a distributed lock in MailReCache while fetching recs from Runtime |
We have introduced a Hazelcast-distributed lock to optimize email recommendation retrieval and prevent redundant server requests. Now, the initial user request captures a lock, fetches recommendations from the RRServer, and caches them in Cassandra. Subsequent requests retrieve recommendations directly from Cassandra, improving response times and reducing server load. |
18-Apr-2024 |
|
|
Feed Processing: Support for Storing Additional Product Attributes in Feed |
We have introduced a new method for handling large, infrequently changed product attributes. Clients can now send these attributes in a separate file （for example, catalog_attributes_sitename_date.*.zip）, which will be stored upon receipt and only utilized during the next full catalog update. This change eliminates the need for frequent re-processing of unchanged data, streamlining operations and reducing processing load. |
18-Apr-2024 |
|
|
Configurable Glass Views in User Profile Service
|
To enhance the flexibility of the User Profile Service (UPS), we've added a new configuration option in the Site Configuration. When "Enable Glass Views in UPS" is activated, UPS begins to store Glass Views (Email Views). Additionally, the UPS Reader will return Glass views when the filter is set to "all" or explicitly includes "glassViews". |
18-Apr-2024 |
Bug and Support Fixes
The following issues have been fixed in the release version 24.08 during Apr 05, 2024 - Apr 18, 2024.
|
Jira# |
Title |
Summary |
General Availability |
|---|---|---|---|
|
|
Data Engineering: Sessions Report - Showing Duplicate Orders & Not Using Currency Divisor |
We have resolved a bug in the Sessions Report where orders were not being deduplicated and the currency divisor was not applied correctly. These fixes ensure the accuracy and reliability of financial data within the report, enhancing its utility for business analysis. |
18-Apr-2024 |
|
|
Enterprise Dashboard, Find: Search Term Expansion - Pagination incorrect when filtering on rules with a search |
We have resolved a bug affecting pagination in the Sessions Report when filtering with a search term, such as "iPhone." Previously, clicking to the next page of filtered results incorrectly displayed the same items as the first page. This was caused by improper API calls that did not include the necessary offset parameters. The corrected implementation now adjusts the API endpoint to properly handle pagination parameters ("/api/productsearch/synonyms?offset=25&rows=25&kw=iphone"), ensuring that each page displays the correct set of items. |
18-Apr-2024 |
|
|
Recommend: Critical Severity Findings (SQL Injection) |
We have fixed a critical SQL injection vulnerability that arose from improperly constructed SQL queries dynamically concatenated with user-supplied data. By transitioning to parameterized queries, we've fortified our database interactions, preventing attackers from manipulating SQL commands to access or alter data. This enhancement ensures the integrity and security of our database operations across all affected endpoints. |
18-Apr-2024 |
|
|
Find: Search term expansion export and pagination issues when filtering on rules with search |
We resolved two significant issues on the Search Term Expansions page: Duplicates in Export: Fixed an issue where exported synonyms showed duplicates not present on the portal, ensuring exports now correctly reflect unique portal entries. Pagination Errors During Searches: Corrected pagination errors when filtering by search terms (e.g., "iPhone"), where navigating to subsequent pages incorrectly repeated items. We improved API calls to properly handle offsets, ensuring accurate page-to-page navigation. |
18-Apr-2024 |
|
|
Streaming Catalog: Few updates are not persisted in Itemstore |
We've resolved an issue in the streaming itemstore where subsequent updates within a one-minute interval mistakenly overwrote previous entries due to a caching mechanism in the MapDB 3 KV library. By switching to a cacheless ValueReader.direct for data writes, we now ensure all updates are preserved accurately. |
18-Apr-2024 |
|
|
Enterprise Dashboard, Social Message: Social Proof - Tracking messages showing multiple messages even though it is not enabled |
We have fixed an issue where Social Proof tracking incorrectly displayed multiple messages even though the feature was disabled. Now, the trackExperience API will correctly show only a single message when multiple messages are not enabled. |
18-Apr-2024 |
|
|
Data Engineering, Enterprise Dashboard: Social Proof - Track Experience Event API call logging both False and Error |
We have addressed an issue with the Social Proof Track Experience Event API where redundant calls were made showing both 'False' and 'Error' statuses for SP Message Status within milliseconds for the same user. This occurred under the new template settings when no message was actually displayed. This redundancy in logging has been eliminated, ensuring that the API now triggers only the necessary call depending on the actual message status. |
18-Apr-2024 |
|
|
Advanced Merchandising: Issue with System Filters and Backfill in Advanced Merch Rules |
We resolved a problem, where using the "Enable system filters and rules" option in Advanced Merchandising rules caused issues when the "Backfill remaining slots with recommendations" option was also employed. |
18-Apr-2024 |
|
|
Streaming Catalog: Purging old, archived dataset during publishing causing Cassandra timeout |
We have addressed the issue, where purging old archived datasets during publishing caused Cassandra timeouts. The system now effectively manages dataset archives without exceeding the limit of three, ensuring stability and preventing timeouts during data purging. |
18-Apr-2024 |