Virtual agent for improving item identification using natural language processing and machine learning techniques
A first device may receive, from a second device, an input corresponding to a search for an item, may identify a set of items, and may determine a set of trend scores associated with the set of items. The first device may determine a set of sentiment scores, and may identify a subset of items, of the set of items, based on the set of trend scores and the set of sentiment scores. The first device may provide, to the second device, information associated with the subset of items to permit the second device to provide, for display, the information associated with the subset of items, and may receive, from the second device, information associated with a selected item, of the subset of items. The first device may determine a return score associated with the selected item, and may perform an action based on the return score.
1. A first device, comprising:
one or more processors to:
receive, from a second device associated with a virtual agent application, an input corresponding to a search for an item,
the second device being associated with a user;
identify a set of items based on receiving the input corresponding to the search for the item;
determine a set of trend scores associated with the set of items based on identifying the set of items,
the set of trend scores being indicative of respective popularities of items, of the set of items, across a time frame, and
the set of trend scores being determined based on a first set of data associated with a first data source;
determine a set of sentiment scores associated with the set of items based on identifying the set of items,
the set of sentiment scores being indicative of respective sentiments of the user towards the set of items, and
the set of sentiment scores being determined based on a second set of data associated with a second data source that is different than the first data source;
identify a subset of items, of the set of items, based on the set of trend scores and the set of sentiment scores;
provide, to the second device, information associated with the subset of items to permit the second device to provide, for display, the information associated with the subset of items;
receive, from the second device, information associated with a selected item, of the subset of items, based on providing the information associated with the subset of items;
determine a return score associated with the selected item based on receiving the information associated with the selected item,
the return score being indicative of a probability of the selected item being returned to an entity after being acquired from the entity, and
the return score being determined based on a third set of data associated with a third data source that is different than the first data source and the second data source; and
perform an action based on the return score.
2. The first device of claim 1 , where the one or more processors are further to:
determine a set of similarity scores associated with the item and the set of items; and
where the one or more processors, when identifying the set of items, are to:
identify the set of items based on the set of similarity scores associated with the item and the set of items.
3. The first device of claim 1 , where the one or more processors are further to:
receive information associated with a purchase history of the user; and
where the one or more processors, when determining the set of sentiment scores, are to:
determine the set of sentiment scores based on the information associated with the purchase history of the user.
4. The first device of claim 1 , where the one or more processors are further to:
receive information associated with social media activity of the user; and
where the one or more processors, when determining the set of sentiment scores, are to:
determine the set of sentiment scores based on the information associated with the social media activity of the user.
5. The first device of claim 1 , where the one or more processors are further to:
receive information that identifies a number of sales of the selected item;
receive information that identifies a number of returns of the selected item; and
where the one or more processors, when determining the return score associated with the selected item, are to:
determine the return score based on the information that identifies the number of sales of the item and the information that identifies the number of returns of the selected item.
6. The first device of claim 1 , where the one or more processors are further to:
perform a natural language processing technique using the input corresponding to the search for the item; and
where the one or more processors, when identifying the set of items, are to:
identify the set of items based on performing the natural language processing technique using the input corresponding to the search for the item.
7. The first device of claim 1 , where the first set of data is associated with another entity that is different from the entity.
8. A method, comprising:
receiving, by a first device and from a second device associated with a virtual agent application, an input corresponding to a search for an item,
the second device being associated with a user;
identifying, by the first device, a set of items based on receiving the input corresponding to the search for the item;
determining, by the first device, a set of trend scores associated with the set of items based on identifying the set of items,
the set of trend scores being indicative of respective popularities of items, of the set of items, across a time frame;
determining, by the first device, a set of sentiment scores associated with the set of items based on identifying the set of items,
the set of sentiment scores being indicative of respective sentiments of the user towards the items of the set of items;
identifying, by the first device, a subset of items, of the set of items, based on the set of trend scores and the set of sentiment scores;
providing, by the first device and to the second device, information associated with the subset of items to permit the second device to provide, for display, the information associated with the subset of items;
receiving, by the first device and from the second device, information associated with a selected item, of the subset of items, based on providing the information associated with the subset of items;
determining, by the first device, a return score associated with the selected item based on receiving the information associated with the selected item,
the return score being indicative of a probability of the selected item being returned to an entity after being acquired from the entity; and
performing, by the first device, an action based on the return score.
9. The method of claim 8 , further comprising:
determining a set of similarity scores associated with a description of the item and a set of other descriptions of the set of items; and
where identifying the set of items comprises:
identifying the set of items based on the set of similarity scores.
10. The method of claim 8 , further comprising:
receiving information associated with a browsing history of the user; and
where determining the set of sentiment scores comprises:
determining the set of sentiment scores based on the information associated with the browsing history of the user.
11. The method of claim 8 , further comprising:
receiving information associated with social media activity of the user; and
where determining the set of sentiment scores comprises:
determining the set of sentiment scores based on the information associated with the social media activity of the user.
12. The method of claim 8 , further comprising:
receiving information associated with a set of values of the selected item; and
where determining the return score comprises:
determining the return score based on the information associated with the set of values of the selected item.
13. The method of claim 8 , further comprising:
receiving information associated with a number of returns of the selected item; and
where determining the return score comprises:
determining the return score based on the information associated with the number of returns of the selected item.
14. The method of claim 8 , further comprising:
receiving information that identifies a number of sales of the set of items; and
where determining the set of trend scores comprises:
determining the set of trend scores based on the information that identifies the number of sales of the set of items.
15. A non-transitory computer-readable medium storing instructions, the instructions comprising:
one or more instructions that, when executed by one or more processors, cause the one or more processors to:
receive, from a device associated with a virtual agent application, an input corresponding to a search for an item,
the device being associated with a user;
identify a set of items based on receiving the input corresponding to the search for the item;
determine a set of trend scores associated with the set of items based on identifying the set of items,
the set of trend scores being indicative of respective popularities of items, of the set of items, across a time frame, and
the set of trend scores being determined based on a first set of data associated with a first data source;
determine a set of sentiment scores associated with the set of items based on identifying the set of items,
the set of sentiment scores being indicative of respective sentiments of the user towards the items of the set of items, and
the set of sentiment scores being determined based on a second set of data associated with a second data source that is different than the first data source;
identify a subset of items, of the set of items, based on the set of trend scores and the set of sentiment scores;
provide, to the device, information associated with the subset of items to permit the device to provide, for display, the information associated with the subset of items;
receive, from the device, information associated with a selected item, of the subset of items, based on providing the information associated with the subset of items;
determine a return score associated with the selected item based on receiving the information associated with the selected item,
the return score being indicative of a probability of the selected item being returned to an entity after being acquired from the entity, and
the return score being determined based on a third set of data associated with a third data source that is different than the first data source and the second data source; and
perform an action based on the return score.
16. The non-transitory computer-readable medium of claim 15 , where the one or more instructions, when executed by the one or more processors, further cause the one or more processors to:
receive text associated with the input corresponding to the search for the item;
determine a set of similarity scores associated with the text and the set of items; and
where the one or more instructions, that cause the one or more processors to identify the set of items, cause the one or more processors to:
identify the set of items based on the set of similarity scores.
17. The non-transitory computer-readable medium of claim 15 , where the one or more instructions, when executed by the one or more processors, further cause the one or more processors to:
receive information associated with a transaction history between the user and the entity; and
where the one or more instructions, that cause the one or more processors to determine the return score associated with the selected item, cause the one or more processors to:
determine the return score based on the information associated with the transaction history between the user and the entity.
18. The non-transitory computer-readable medium of claim 15 , where the one or more instructions, when executed by the one or more processors, further cause the one or more processors to:
receive information associated with social media activity of the user; and
where the one or more instructions, that cause the one or more processors to determine the set of sentiment scores, cause the one or more processors to:
determine the set of sentiment scores based on the information associated with the social media activity of the user.
19. The non-transitory computer-readable medium of claim 15 , where the action corresponds to adjusting a value associated with the selected item.
20. The non-transitory computer-readable medium of claim 15 , where the one or more instructions, when executed by the one or more processors, further cause the one or more processors to:
receive information associated with a set of other entities,
the other entities being different than the entity; and
where the one or more instructions, that cause the one or more processors to determine the set of trend scores, cause the one or more processors to:
determine the set of trend scores based on the information associated with the set of other entities.