PREDICTING WHEN A USER IS IN NEED OF A LOAN AND NOTIFYING THE USER OF LOAN OFFERS
Techniques are disclosed to determine when a user is in need of a loan and notifying the user of loan offers. With user permission or affirmative consent, user data may be monitored for several users, which is used to build a user profile for each user. The user profile may then be analyzed to determine whether a user will require a loan within a future time period. To do so, the user data may include data from various sources, which indicate the user's interactions and behaviors such as demographic data, data indicative of user shopping habits, online browsing, life events, or other relevant behaviors. This data may then be analyzed to predict a statistical likelihood that a user will need a loan. When this statistical likelihood is exceeded, a user may be preapproved for a loan and/or a targeted notification may be sent indicating offers for certain types of loans.
1 . A computer-implemented method comprising:
training, by one or more processors, a machine learning model to determine a set of weights based on historical profiles associated with a set of users, wherein the set of weights indicate correlations between:
a set of user inputs indicated in the historical profiles, and
instances of individual users, in the set of users, obtaining loans;
receiving, by the one or more processors, user input data associated with a user;
generating, by the one or more processors, a user profile based upon the user input data;
predicting, by the one or more processors and using the machine learning model, a statistical likelihood that the user will require a loan within a future time period, by:
identifying at least one user input, of the set of user inputs, indicated in the user profile;
determining at least one weight, of the set of weights, associated with the at least one user input; and
determining the statistical likelihood based on the at least one weight;
determining, by the one or more processors, that the statistical likelihood exceeds a threshold likelihood; and
based at least in part on the statistical likelihood exceeding the threshold likelihood, transmitting, by the one or more processors, a notification to a computing device associated with the user, wherein the notification includes one or more offers for one or more customized loans.
2 . The computer-implemented method of claim 1 , further comprising:
calculating, by the one or more processors and based at least in part on the user profile, a set of probabilities associated with a range of loan amounts, wherein probabilities in the set of probabilities indicate respective likelihoods that the loan will be for corresponding loan amounts;
identifying, by the one or more processors, a loan amount associated with a highest probability of the set of probabilities; and
preapproving, by the one or more processors, the user for the loan amount.
3 . The computer-implemented method of claim 1 , further comprising:
receiving, by the one or more processors as part of the user input data, a web browsing history associated with the user;
identifying, by the one or more processors, one or more of search terms and websites from the web browsing history that are relevant to the user; and
storing, by the one or more processors, the one or more of the search terms and websites as part of the user profile,
wherein the statistical likelihood is based at least in part upon the one or more of the search terms and websites.
4 . The computer-implemented method of claim 1 , wherein the user input data indicates a current age of a family member of the user, the method further comprising:
determining, by the one or more processors and based at least in part on the current age, a number of days until the family member reaches a legal driving age; and
determining, by the one or more processors, that the statistical likelihood exceeds the threshold likelihood based at least in part on the number of days being less than a threshold number of days.
5 . The computer-implemented method of claim 1 , wherein the user input data indicates one or more life events associated with the user.
6 . The computer-implemented method of claim 1 , wherein the notification is one or more of:
(i) a text message,
(ii) an email message, or
(iii) a push notification.
7 . (canceled)
8 . A computer system comprising:
one or more processors;
memory storing computer-executable instructions that, when executed by the one or more processors, cause the one or more processors to:
train a machine learning model to determine a set of weights based on historical profiles associated with a set of users, wherein the set of weights indicates correlations between:
a set of user inputs indicated in the historical profiles, and
instances of individual users, in the set of users, obtaining loans;
receive user input data from a client device associated with a user;
generate a user profile based upon the user input data;
predict, using the machine learning model, a statistical likelihood that the user will require a loan within a future time period, by
identifying at least one user input, of the set of user inputs, indicated in the user profile;
determining at least one weight, of the set of weights, associated with the at least one user input; and
determining the statistical likelihood based on the at least one weight;
determine that the statistical likelihood exceeds a threshold likelihood; and
based at least in part on the statistical likelihood exceeding the threshold likelihood, transmit a notification to the client device, wherein the notification includes one or more offers for one or more customized loans.
9 . The computer system of claim 8 , wherein the computer-executable instructions further cause the one or more processors to:
calculate, based at least in part on the user profile, a set of probabilities associated with a range of loan amounts, wherein probabilities in the set of probabilities indicate respective likelihoods that the loan will be for corresponding loan amounts;
identify a loan amount associated with a highest probability of the set of probabilities; and
preapprove the user for the loan amount.
10 . The computer system of claim 8 , wherein:
the user input data includes a web browsing history associated with the user,
the computer-executable instructions further cause the one or more processors to:
(i) identify one or more of search terms and websites from the web browsing history that are relevant to the user requiring the loan, and
(ii) store the one or more of the search terms and websites as part of the user profile, and
the statistical likelihood is based at least in part upon the one or more of the search terms and websites.
11 . The computer system of claim 8 , wherein the user input data indicates a current age of a family member of the user, and the computer-executable instructions further cause the one or more processors to:
(i) determine, based at least in part on the current age, a number of days until the family member reaches a legal driving age, and
(ii) determine that the statistical likelihood exceeds the threshold likelihood based at least in part on when the number of days being less than a threshold number of days.
12 . The computer system of claim 8 , wherein the user input data indicates one or more life events associated with the user.
13 . The computer system of claim 8 , wherein the notification is one or more of:
(i) a text message,
(ii) an email message, or
(iii) a push notification.
14 . (canceled)
15 . A non-transitory computer-readable medium storing instructions thereon that, when executed by one or more processors, cause the one or more processors to perform operations comprising:
training a machine learning model to determine a set of weights based on historical profiles associated with a set of users, wherein the set of weights indicate correlations between:
a set of user inputs indicated in the historical profiles, and
instances of individual users, in the set of users, obtaining loans;
receiving user input data associated with a user from a mobile device of the user;
generating a user profile based upon the user input data;
predicting, using the machine learning model, a statistical likelihood that the user will require a loan within a future time period, by:
identifying at least one user input, of the set of user inputs, indicated in the user profile;
determining at least one weight, of the set of weights, associated with the at least one user input; and
determining the statistical likelihood based on the at least one weight;
determining that the statistical likelihood exceeds a threshold likelihood; and
based at least in part on the statistical likelihood exceeding the threshold likelihood, transmitting a notification to the mobile device,
wherein the notification includes one or more offers for one or more customized loans.
16 . The non-transitory computer-readable medium of claim 15 , wherein the operations further comprise:
receiving, as part of the user input data, a web browsing history associated with the user;
identifying one or more of search terms and web sites from the web browsing history that are relevant to the user requiring the loan; and
storing the one or more of the search terms and web sites as part of the user profile,
wherein the statistical likelihood is based at least in part upon the one or more of the search terms and websites.
17 . The non-transitory computer-readable medium of claim 15 , wherein the user input data indicates a current age of a family member of the user, and the operations further comprise:
determining, based at least in part on the current age, a number of days until the family member reaches a legal driving age; and
determining that the statistical likelihood exceeds the threshold likelihood based at least in part on the number of days being less than a threshold number of days.
18 . The non-transitory computer-readable medium of claim 15 , wherein the user input data indicates one or more life events associated with the user.
19 . (canceled)
20 . (canceled)
21 . The computer-implemented method of claim 1 , wherein the computing device associated with the user is a mobile device, and the one or more processors receives the user input data from the mobile device.
22 . The computer-implemented method of claim 1 , further comprising determining, by the one or more processors, at least one of a loan type, a loan term, and a monetary amount associated with the one or more customized loans.
23 . The computer-implemented method of claim 1 , wherein the threshold likelihood comprises a first threshold likelihood, the method further comprising:
determining, by the one or more processors, that the statistical likelihood exceeds a second threshold likelihood, the second threshold likelihood being higher than the first threshold likelihood; and
pre-approving, by the one or more processors, the user for the one or more customized loans based on determining that the statistical likelihood exceeds the second threshold likelihood.
24 . The computer-implemented method of claim 1 , further comprising:
determining, by the one or more processors, that the at least one user input corresponds to a predetermined condition associated with a particular type of loan; and
determining, by the one or more processors, that the statistical likelihood exceeds the threshold likelihood, based on determining that the at least one user input corresponds to the predetermined condition,
wherein the notification includes an offer associated with the particular type of loan.