System and method for online analysis

A method for automated analysis using machine learning, comprising:

• receiving, at a server computer from an authorizing entity computer, a request to dispute a completed transaction of a user of a user computer, wherein the request includes a device identifier of the user computer;
• identifying, by the server computer, credentials used in connection with a plurality of completed transactions based on the device identifier of the user computer, wherein the server computer stores a mapping between the device identifier and the credentials;
• gathering in real-time, by the server computer from a database accessible by the server computer, additional information about the plurality of completed transactions based on the credentials, wherein the additional information includes an attribute associated with an item or a service provided in the plurality of completed transactions;
• identifying in real-time, by the server computer among the plurality of completed transactions, an individual completed transaction conducted between the user and a resource provider based on the additional information, wherein the server computer is a third party with respect to the user and the resource provider;
• initiating, by the server computer, an online session with the user computer based on the request and the additional information;
• presenting, by the server computer, the individual completed transaction to the user computer as part of the online session;
• receiving, by the server computer, a user input confirming the individual completed transaction as a selected completed transaction as part of the online session;
• determining, by the server computer and from a database, a set of attributes including one or more attributes associated with the selected completed transaction and one or more attributes associated with the resource provider;
• identifying, by the server computer, a ruleset applicable to the selected completed transaction at a ruleset database based on at least the set of attributes;
• generating in real-time, by the server computer using a machine learning algorithm, a first question pertaining to the selected completed transaction based on the set of attributes and the ruleset applicable to the selected completed transaction;
• presenting, by the server computer, the first question pertaining to the selected completed transaction to the user computer as part of the online session;
• receiving, from the user computer and by the server computer, a first response to the first question as part of the online session;
• generating, by the server computer using the machine learning algorithm, a second question based on the set of attributes, the ruleset applicable to the selected completed transaction and the first response;
• presenting, by the server computer, the second question pertaining to the selected completed transaction to the user computer as part of the online session;
• receiving, from the user computer and by the server computer, a second response to the second question as part of the online session;
• and
• storing, by the server computer, the received first response and the received second response in a data storage element, wherein the data storage element is accessible by the authorizing entity computer that issued the credentials used in the individual completed transaction between the user and the resource provider.