Automatic generation of featured filters

1. A system comprising:

one or more processors; and

one or more non-transitory computer-readable media storing computing instructions configured to run on the one or more processors and perform:

selectively aggregating a first set of filters for an item results list, the first set of filters comprising multiple filter types, the multiple filter types of the first set of filters comprising category filters, facet filters, and price filters;

for each filter in the first set of filters and each item in the item results list, determining user engagement statistics for the item when the filter has been applied;

generating a filter score for each filter in the first set of filters;

selecting a second set of filters from the first set of filters based on the filter scores of the filters in the second set of filters being above a threshold filter score, wherein the threshold filter score is a predetermined percentage of a highest filter score from among the filter scores for the filters in the first set of filters;

applying redundancy rules to remove redundant filters from the second set of filters;

applying space-constraint rules to the second set of filters to limit a quantity of filters in the second set of filters based on a ranking of the filters in the second set of filters and based on a screen size of a user device;

applying image-selection rules to select a respective display image for each of the filters in the second set of filters;

applying mutual-information rules to determine whether a mutual-information score for the second set of filters exceeds a predetermined mutual-information threshold; and

after receiving a request to display a webpage that lists at least a portion of the item results list, coordinating a display of the webpage, wherein the webpage comprises the respective display image for each of the filters in the second set of filters and a set of options to select any of the filters in the second set of filters when the mutual-information score for the second set of filters exceeds the predetermined mutual-information threshold.

2. The system of claim 1 , wherein the computing instructions are further configured to perform:

after selecting the second set of filters and before applying the space-constraint rules, applying de-scattering rules to group the filters in the second set of filters.

3. The system of claim 1 , wherein the computing instructions are further configured to perform:

after selecting the second set of filters and before applying the space-constraint rules, applying re-ranking rules to reorder the filters in the second set of filters.

4. The system of claim 1 , wherein:

the item results list is a search-query item results list;

the first set of filters further comprise one or more query reformulations; and

the request to display the webpage is a search query from a user operating the user device.

5. The system of claim 1 , wherein:

the item results list is a category-browsing item results list; and

the request to display the webpage is a category browse request from a user operating the user device.

6. The system of claim 1 , wherein the computing instructions are further configured to perform:

applying name-generation rules to generate a display name for each filter in the first set of filters based on a type of the filter.

7. The system of claim 1 , wherein:

the user engagement statistics comprise item impressions when the filter is applied, item clicks when the filter is applied, item add-to-carts when the filter is applied, and item orders when the filter is applied.

8. The system of claim 1 , wherein:

the user engagement statistics are based on a previous at least 90 days of historical user logs.

9. The system of claim 1 , wherein:

generating the filter score for each filter in the first set of filters comprises:

generating an item-filter score for each item in the item results list for the filter; and

aggregating the item-filter scores across all items in the item result list for the filter.

10. A method being implemented by one or more processors the method comprising:

selectively aggregating a first set of filters for an item results list, the first set of filters comprising multiple filter types, the multiple filter types of the first set of filters comprising category filters, facet filters, and price filters, each filter of the first set of filters having been applied to the item results list;

for each filter in the first set of filters and each item in the item results list, determining user engagement statistics for the item when the filter has been applied;

generating a filter score for each filter in the first set of filters;

selecting a second set of filters from the first set of filters based on the filter scores of the filters in the second set of filters being above a threshold filter score, wherein the threshold filter score is a predetermined percentage of a highest filter score from among the filter scores for the filters in the first set of filters;

applying redundancy rules to remove redundant filters from the second set of filters;

applying space-constraint rules to the second set of filters to limit a quantity of filters in the second set of filters based on a ranking of the filters in the second set of filters and based on a screen size of a user device;

applying image-selection rules to select a respective display image for each of the filters in the second set of filters;

applying mutual-information rules to determine whether a mutual-information score for the second set of filters exceeds a predetermined mutual-information threshold;

determining that the mutual-information score exceeds the predetermined mutual-information threshold; and

after receiving a request to display a webpage that lists at least a portion of the item results list and based on determining that the mutual-information score exceeds the predetermined mutual-information threshold, coordinating a display of the webpage, wherein the webpage comprises the respective display image for each of the filters in the second set of filters and a set of options to select any of the filters in the second set of filters.

11. The method of claim 10 further comprising:

after selecting the second set of filters and before applying the space-constraint rules:

applying de-scattering rules to group the filters in the second set of filters.

12. The method of claim 10 , wherein:

the item results list is a search-query item results list;

the first set of filters further comprise one or more query reformulations; and

the request to display the webpage is a search query from a user operating the user device.

13. The method of claim 10 , wherein:

the item results list is a category-browsing item results list; and

the request to display the webpage is a category browse request from a user operating the user device.

14. The method of claim 10 , further comprising:

applying name-generation rules to generate a display name for each filter in the first set of filters based on a type of the filter.

15. The method of claim 10 , wherein:

the user engagement statistics comprise item impressions when the filter is applied, item clicks when the filter is applied, item add-to-carts when the filter is applied, and item orders when the filter is applied.

16. The method of claim 10 , wherein:

generating the filter score for each filter in the first set of filters comprises:

generating an item-filter score for each item in the item results list for the filter; and

aggregating the item-filter scores across all items in the item result list for the filter.

17. The method of claim 10 , wherein:

after selecting the second set of filters and before applying the space-constraint rules:

applying re-ranking rules to reorder the filters in the second set of filters.

18. The method of claim 10 , wherein:

the user engagement statistics are based on a previous at least 90 days of historical user logs.