Amazon Personalize has launched a new integration with Amazon OpenSearch Service that allows you to personalize search results for each user and helps predict their search needs. The Amazon Personalize Search Ranking plugin within OpenSearch Service allows you to improve end-user engagement and conversion from your website and app search by leveraging the deep learning capabilities offered by Amazon Personalize. This feature is also available with self-managed OpenSearch.
Search is crucial to engaging users because it generates high-intent traffic from people searching for specific products or categories. Previously, clients found it difficult to capitalize on this traffic and provide relevant search results to their users due to infrastructure limitations or lack of ML expertise. This caused an increase in cases of users not finding the items they were looking for. With the Amazon Personalize Search Ranking plugin, OpenSearch Service version 2.9.0 or later customers can go beyond the traditional keyword matching approach and enhance relevant elements in an individual user’s search results based on their interests, context and past interactions in real time. You can also adjust the level of customization for each search query to ensure flexibility and control over the search experience.
AWS Partners Like Competent are excited about the personalization possibilities that the Amazon Personalize Search Ranking plugin will unlock for their retail and media clients.
“Amazon Personalize has been proven to be highly impactful for many businesses with its cost-effective and streamlined implementation. With the release of the new Amazon Personalize Search Ranking plugin within Amazon OpenSearch Service, we can now quickly implement and implement real-time user personalization for search results. We are very confident that it will provide a better customer experience and satisfaction, as well as increasing conversion and click rates by two to three times. Personalized search is a differentiator, especially for media and retail platforms. “We are very excited to be a launch partner for AWS on this release and look forward to helping businesses deliver personalized search solutions powered by Amazon Personalize.”
– Andy Huang, Head of ai/ML at Cognizant Servian.
In this post, we show you how search results are customized by user and how they vary when you adjust the personalization weight. We specify a value closer to zero to place less emphasis on personalization, and we specify a value closer to 1 to reclassify search results based on a higher level of personalization.
Example use cases
To explore the impact of this new feature in more detail, let’s review an example using a data set from Retail Demo Store.
First, we use the OpenSearch service to get search results for the search query “Grooming.” When the customization weight is set to 0.0, no customization is performed. As shown in the table below, the top five search results from the OpenSearch service show grooming items with a higher gender affinity toward women (see Gender_Affinity
column, where M
means masculine and F
means woman).
Range | Article ID | Article name | Description | Gender_affinity |
1 | 1bcb66c4-ee9d-4c0c-ba53-168cb243569f | Women’s grooming kit | Essential in every bathroom | F |
2 | f91ec34f-a08e-4408-8bb0-592bdd09375c | Besto Women’s Hair Brush | Soft brush for daily use | F |
3 | 4296626c-fbb0-42b4-9a50-b6c6c16095f3 | Makeup Brushes Kit | This ingenious makeup brush kit is essential in every… | F |
4 | 09920b2e-4e07-41f7-aca6-47744777a2a7 | Fashion razor | Essential in every bathroom | F |
5 | 39945ad0-57c9-4c28-a69c-532d5d167202 | Make up brushes | Makeup brushes for every bathroom. | F |
6 | 1bfbe5c7-6f02-4465-82f1-6083a4b302c0 | Premium men’s razor | Razor for every bathroom | METER |
7 | 6d5b3f03-ade6-42f7-969d-acd1f2162332 | 5 blade razor for men | Razor for every bathroom | METER |
8 | 83095a08-2968-4275-a375-4fab404df7ac | Razers Fusion5 Men’s | Razor for every bathroom | METER |
9 | afdd9c41-2762-45bf-b6a7-e3fb8f1b34ba | Minimalist knife | Essential in every bathroom | METER |
10 | 5dbc7cb7-39c5-4795-9064-d1655d78b3ca | Razor brand for men | Razor for every bathroom | METER |
Suppose a user with gender M
(male) performs a search using the same query for “Grooming.” When the personalization weight is set to 0.3, items with gender affinity towards men get a subtle increase in ranking. In this example, Premium Men’s Razor, which was originally ranked #6 in the previous table according to OpenSearch Service, moves to #2 in the updated table. Similarly, Razor Brand for Men appears higher (ranked 6) despite being the lowest ranked item in the table above.
Range | Article ID | Article name | Description | Gender_affinity |
1 | 1bcb66c4-ee9d-4c0c-ba53-168cb243569f | Women’s grooming kit | Essential in every bathroom | F |
2 | 1bfbe5c7-6f02-4465-82f1-6083a4b302c0 | Premium men’s razor | Razor for every bathroom | METER |
3 | f91ec34f-a08e-4408-8bb0-592bdd09375c | Besto Women’s Hair Brush | Soft brush for daily use | F |
4 | 4296626c-fbb0-42b4-9a50-b6c6c16095f3 | Makeup Brushes Kit | This ingenious makeup brush kit is essential in every… | F |
5 | 09920b2e-4e07-41f7-aca6-47744777a2a7 | Fashion razor | Essential in every bathroom | F |
6 | 5dbc7cb7-39c5-4795-9064-d1655d78b3ca | Razor brand for men | Razor for every bathroom | METER |
7 | 39945ad0-57c9-4c28-a69c-532d5d167202 | Make up brushes | Makeup brushes for every bathroom. | F |
8 | afdd9c41-2762-45bf-b6a7-e3fb8f1b34ba | Minimalist knife | Essential in every bathroom | METER |
9 | 83095a08-2968-4275-a375-4fab404df7ac | Razers Fusion5 Men’s | Razor for every bathroom | METER |
10 | 6d5b3f03-ade6-42f7-969d-acd1f2162332 | 5 blade razor for men | Razor for every bathroom | METER |
Next, we set the personalization weight to a value of 0.8 to get more personalized search results for “Careful.” In the table below, the top four elements of the search results are highly suitable for men. Premium men’s razors and men’s razor brands move up the ranks even further. We also see that other grooming items, such as the Minimalistic Razor and Fusion5 Razers for Men, appeared at the top of the search results even though they ranked lower in our first query.
Range | Article ID | Article name | Description | Gender_affinity |
1 | 1bfbe5c7-6f02-4465-82f1-6083a4b302c0 | Premium men’s razor | Razor for every bathroom | METER |
2 | 5dbc7cb7-39c5-4795-9064-d1655d78b3ca | Razor brand for men | Razor for every bathroom | METER |
3 | afdd9c41-2762-45bf-b6a7-e3fb8f1b34ba | Minimalist knife | Essential in every bathroom | METER |
4 | 83095a08-2968-4275-a375-4fab404df7ac | Razers Fusion5 Men’s | Razor for every bathroom | METER |
5 | 1bcb66c4-ee9d-4c0c-ba53-168cb243569f | Women’s grooming kit | Essential in every bathroom | F |
6 | f91ec34f-a08e-4408-8bb0-592bdd09375c | Besto Women’s Hair Brush | Soft brush for daily use | F |
7 | 6d5b3f03-ade6-42f7-969d-acd1f2162332 | 5 blade razor for men | Razor for every bathroom | METER |
8 | 09920b2e-4e07-41f7-aca6-47744777a2a7 | Fashion razor | Essential in every bathroom | F |
9 | 39945ad0-57c9-4c28-a69c-532d5d167202 | Make up brushes | Makeup brushes for every bathroom. | F |
10 | 4296626c-fbb0-42b4-9a50-b6c6c16095f3 | Makeup Brushes Kit | This ingenious makeup brush kit is essential in every… | F |
For more details on how to implement custom search with the OpenSearch service, see Customizing OpenSearch search results.
Conclusion
With the new Amazon Personalize Search Ranking plugin, self-managed OpenSearch and OpenSearch Service v2.9 and higher customers can enhance relevant elements in their search results by including signals from each user’s history, context, and preferences. The plugin allows you to exercise greater control over the level of customization for each user and query type, and improve the overall search experience for your users.
For more details about Amazon Personalize, see the Amazon Personalize Developer Guide.
About the authors
Sharma Sharma is a Senior Technical Product Manager working with AWS ai/ML on the Amazon Personalize team. He has experience in computer engineering, technology consulting and data analysis.
Kulkarni Sticky Rice is a software development engineer on the Amazon Personalize team focused on building ai-based recommendation systems at scale. In his free time he likes to read and travel.
Prashant Mishra is a software development engineer on the Amazon Personalize team.
Branislav Kveton is a Principal Scientist at AWS ai Labs. He proposes, analyzes, and applies algorithms that learn incrementally, run in real time, and converge toward near-optimal solutions as the number of observations increases.