Method and system for generating real-time personalized advertisements in data management self-help systems

1. A system comprising:

at least one processor; and

at least one memory storing instructions that, when executed by the at least one processor, cause the system to:

train, with one or more machine learning processes, a predictive model to match assistance documents from an assistance document database of a financial management system to messaging content stored in a messaging content database, wherein the predictive model includes at least a Latent Dirichlet Allocation model and an uplift model;

receive user related data including at least demographic data related to a user and financial data related to the user;

receive a search query from the user for information regarding one or more topics associated with the financial management system;

receive, based on the search query, an access request from the user to access an assistance document from the assistance document database;

analyze the assistance document with the predictive model;

generate messaging topic matching data with the predictive model by matching the assistance document to messaging content from the messaging content database based on analysis of the assistance document and the user related data;

generate a personalized message based on the messaging topic matching data and rewording the messaging content to mimic a writing style of the user as indicated by the search query, and further based on adjusting the messaging content to include at least a portion of wording included in the assistance document, wherein the personalized message promotes a product or service of the financial management system; and

present the personalized message and the assistance document to the user in real time on a display screen.

2. The system of claim 1 , wherein the Latent Dirichlet Allocation model is an unsupervised Latent Dirichlet Allocation model.

3. The system of claim 1 , wherein the predictive model includes a multi-class classifier.

4. The system of claim 3 , wherein the multi-class classifier is a supervised multi-class classifier.

5. The system of claim 1 , wherein the financial management system includes one or more of a tax return preparation system, a bookkeeping system, an accounting system, a budgeting system, or a financial transaction tracking system.

6. The system of claim 1 , wherein execution of the instructions further causes the processor to:

generate search results based on the search query and including links to one or more assistance documents from the assistance documents database; and

output the search results to the user.

7. The system of claim 6 , wherein receiving the access request includes receiving an indication that the user has selected the assistance document from among the search results.

8. The system of claim 7 , wherein execution of the instructions further causes the system to:

analyze the search query with the predictive model; and

generate the messaging topic matching data based on analysis of the search query.

9. The system of claim 7 , wherein execution of the instructions further causes the system to:

analyze the search query; and

generate the personalized message by adjusting the messaging content to include wording specific to the search query.

10. The system of claim 1 , wherein the user related data further includes data indicating how the user has previously interacted with the financial management system.

11. The system of claim 1 , wherein the personalized message includes an advertisement promoting the product or the service of the financial management system.

12. The system of claim 1 , wherein presenting the personalized message and the assistance document to the user in real time includes presenting the personalized message and the assistance document in a same webpage.

13. The system of claim 1 , wherein receiving the access request includes the user requesting a webpage associated with the assistance document via a web browser.

14. The system of claim 13 , wherein receiving the access request includes the user selecting the webpage from search results provided by a third-party website.

15. A method performed by one or more processors of a system, the method comprising:

training, with one or more machine learning processes, a predictive model to match assistance documents from an assistance document database of a financial management system to messaging content stored in a messaging content database, wherein the predictive model includes at least a Latent Dirichlet Allocation model and an uplift model;

receiving user related data including at least demographic data related to a user and financial data related to the user;

receiving a search query from the user for information regarding one or more topics associated with the financial management system;

receiving, based on the search query, an access request from the user to access an assistance document from the assistance document database;

analyzing the assistance document with the predictive model;

generating messaging topic matching data with the predictive model by matching the assistance document to messaging content from the messaging content database based on analysis of the assistance document and the user related data;

generating a personalized message based on the messaging topic matching data and rewording the messaging content to mimic a writing style of the user as indicated by the search query, and further based on adjusting the messaging content to include at least a portion of wording included in the assistance document, wherein the personalized message promotes a product or service of the financial management system; and

presenting the personalized message and the assistance document to the user in real time on a display screen.

16. The method of claim 15 , wherein the Latent Dirichlet Allocation model is an unsupervised Latent Dirichlet Allocation model.

17. The method of claim 15 , wherein the predictive model includes a multi-class classifier.

18. The method of claim 17 , wherein the multi-class classifier is a supervised multi-class classifier.

19. The method of claim 15 , wherein the financial management system includes one or more of a tax return preparation system, a bookkeeping system, an accounting system, a budgeting system, or a financial transaction tracking system.

20. The method of claim 15 , further comprising:

generating search results based on the search query and including links to one or more assistance documents from the assistance documents database; and

outputting the search results to the user.

21. The method of claim 20 , wherein receiving the access request includes receiving an indication that the user has selected the assistance document from among the search results.

22. The method of claim 21 , further comprising:

analyzing the search query with the predictive model; and

generating the messaging topic matching data based on analysis of the search query.

23. The method of claim 21 , further comprising:

analyzing the search query; and

generating the personalized message by adjusting the messaging content to include wording specific to the search query.

24. The method of claim 15 , wherein the user related data further includes data indicating how the user has previously interacted with the financial management system.

25. The method of claim 15 , wherein the personalized message includes an advertisement promoting the product or the service of the financial management system.

26. The method of claim 15 , wherein presenting the personalized message and the assistance document to the user in real time includes presenting the personalized message and the assistance document in a same webpage.

27. The method of claim 15 , wherein receiving the access request includes the user requesting a webpage associated with the assistance document via a web browser.

28. The method of claim 27 , wherein receiving the access request includes the user selecting the webpage from search results provided by a third-party website.

29. A method performed by one or more processors of a system, the method comprising:

training, with one or more machine learning processes, a predictive model to match a user of a financial management system to messaging content stored in a messaging content database of the financial management system, wherein the predictive model includes at least a Latent Dirichlet Allocation model and an uplift model;

receiving user related data including at least demographic data related to the user and financial data related to the user;

receiving, from the user, a search query for information regarding one or more topics associated with the financial management system;

outputting search results to the user based on the search query;

receiving selection data from the user selecting, from the search results, an assistance document from an assistance document database;

analyzing the search query with the predictive model;

generating messaging topic matching data with the predictive model by matching the user to messaging content from the messaging content database based on analysis of the search query and the user related data;

generating a personalized message based on the messaging topic matching data and rewording the messaging content to mimic a writing style of the user as indicated by the search query, and further based on adjusting the messaging content to include at least a portion of wording included in the assistance document, wherein the personalized message promotes a product or service of the financial management system; and

presenting the personalized message and the assistance document to the user in real time on a display screen.

30. The method of claim 29 , further comprising:

analyzing the assistance document with the predictive model; and

generating the personalized message based on analysis of the assistance document and the search query.