Methods and systems for ranking messaging bots
Exemplary embodiments relate to techniques for identifying messaging robots, or bots, to surface in response to a request. For example, in order to facilitate increased interaction between a user and a bot, a list of candidate bots that the user is likely to be interested in may be surfaced to the user in response to a search for a bot or a request that a bot perform a particular task. Identifying the bots may be accomplished by generating a list of candidate bots and filtering the list based on filtering metrics. Then, the remaining bots may be ranked based on ranking metrics, and the top bots in the ranking may be returned. In some embodiments, two sets of ranks may calculated: one for retention rate, and another rank for the number of messages sent and received by the bots.
1. A method, comprising:
receiving an input relating to a messaging bot request;
identifying a plurality of candidate bots associated with the messaging bot request;
filtering the plurality of candidate bots based on one or more filtering metrics;
ranking the filtered bots based on ranking metrics including at least a retention rate and an engagement rate of the filtered bots;
selecting a predetermined number of top-ranked bots, at least a first portion of the predetermined number being a predefined number of top-ranked bots associated with the retention rate and a second portion of the predetermined number being a predefined number of top-ranked bots associated with the engagement rate; and
returning a list of the selected top-ranked bots in response to the input, and
the retention rate defining a number of users returning to interact with the filtered bots after previous interactions over a predetermined time period, and
the engagement rate defining a measurement of how much or often the filtered bots respond to user queries.
2. The method of claim 1 , wherein the ranking is performed separately for the retention rate and the engagement rate and top bots in each category are returned.
3. The method of claim 1 , at least one metric configured to serve as both a filtering metric and a ranking metric.
4. The method of claim 1 , wherein the input is a request to search for bots, or a task to be performed by a bot.
5. The method of claim 1 , wherein at least one of the candidate bots is associated with a page in a social network, and the filtering metrics include a block rate of the page, a popularity of the page among users of the social network, previous interactions of a user associated with the input with the page, social signals from friends of the user associated with the input, a response rate of the candidate bot, whether the candidate bot is correctly configured, or a language spoken by the candidate bot.
6. The method of claim 1 , wherein at least one of the candidate bots is associated with a page in a social network, and the ranking metrics include the retention rate of the candidate bot, a block rate of the page, a number of messages sent or received by the candidate bot, a popularity of the page among users of the social network, a response rate of the candidate bot, a fan count of the bot or the page, a number of followers of the page, and pixel activity associated with the user and the page.
7. The method of claim 1 , wherein at least one of the ranking metrics is adjusted based on a context, the context including one or more of a category of a page associated with the bot, an age of a user associated with the input, a gender of the user, an interest of the user, a time of day, or a seasonality.
8. A non-transitory computer-readable medium storing instructions configured to cause one or more processors to:
receive an input relating to a messaging bot request;
identify a plurality of candidate bots associated with the messaging bot request;
filter the plurality of candidate bots based on one or more filtering metrics;
rank the filtered bots based on ranking metrics including at least a retention rate and an engagement rate of the filtered bots;
select a predetermined number of top-ranked bots, at least a first portion of the predetermined number being a predefined number of top-ranked bots associated with the retention rate and a second portion of the predetermined number being a predefined number of top-ranked bots associated with the engagement rate; and
return a list of the selected top-ranked bots in response to the input, and
the retention rate defining a number of users returning to interact with the filtered bots after previous interactions over a predetermined time period, and
the engagement rate defining a measurement of how much or often the filtered bots respond to user queries.
9. The computer-readable medium of claim 8 , wherein the ranking is performed separately for the retention rate and the engagement rate and top bots in each category are returned.
10. The computer-readable medium of claim 8 , at least one metric configured to serve as both a filtering metric and a ranking metric.
11. The computer-readable medium of claim 8 , wherein the input is a request to search for bots, or a task to be performed by a bot.
12. The computer-readable medium of claim 8 , wherein at least one of the candidate bots is associated with a page in a social network, and the filtering metrics include a block rate of the page, a popularity of the page among users of the social network, previous interactions of a user associated with the input with the page, social signals from friends of the user associated with the input, a response rate of the candidate bot, whether the candidate bot is correctly configured, or a language spoken by the candidate bot.
13. The computer-readable medium of claim 8 , wherein at least one of the candidate bots is associated with a page in a social network, and the ranking metrics include the retention rate of the candidate bot, a block rate of the page, a number of messages sent or received by the candidate bot, a popularity of the page among users of the social network, a response rate of the candidate bot, a fan count of the bot or the page, a number of followers of the page, and pixel activity associated with the user and the page.
14. The computer-readable medium of claim 8 , wherein at least one of the ranking metrics is adjusted based on a context, the context including one or more of a category of a page associated with the bot, an age of a user associated with the input, a gender of the user, an interest of the user, a time of day, or a seasonality.
15. An apparatus comprising:
a non-transitory computer readable medium storing instructions for performing a bot search; and
a processor configured to execute the instructions to cause the processor to:
receive an input relating to a messaging bot request;
identify a plurality of candidate bots associated with the messaging bot request
filter the plurality of candidate bots based on one or more filtering metrics;
rank the filtered bots based ranking metrics including at least a retention rate and an engagement rate of the filtered bots;
select a predetermined number of top-ranked bots, at least a first portion of the predetermined number being a predefined number of top-ranked bots associated with the retention rate and a second portion of the predetermined number being a predefined number of top-ranked bots associated with the engagement rate; and
return a list of the selected top-ranked bots in response to the input, and
the retention rate defining a number of users returning to interact with the filtered bots after previous interactions over a predetermined time period, and
the engagement rate defining a measurement of how much or often the filtered bots respond to user queries.
16. The apparatus of claim 15 , wherein the ranking is performed separately for the retention rate and the engagement rate and top bots in each category are returned.
17. The apparatus of claim 15 , at least one metric configured to serve as both a filtering metric and a ranking metric.
18. The apparatus of claim 15 , wherein the input is a request to search for bots, or a task to be performed by a bot.
19. The apparatus of claim 15 , wherein at least one of the candidate bots is associated with a page in a social network, and the filtering metrics include a block rate of the page, a popularity of the page among users of the social network, previous interactions of a user associated with the input with the page, social signals from friends of the user associated with the input, a response rate of the candidate bot, whether the candidate bot is correctly configured, or a language spoken by the candidate bot.
20. The apparatus of claim 15 , wherein at least one of the candidate bots is associated with a page in a social network, and the ranking metrics include the retention rate of the candidate bot, a block rate of the page, a number of messages sent or received by the candidate bot, a popularity of the page among users of the social network, a response rate of the candidate bot, a fan count of the bot or the page, a number of followers of the page, and pixel activity associated with the user and the page.