Tax fraud detection through linked relationships

A computerized method for determining a likelihood that a user submitting a subject tax return is genuine, the method comprising the following steps:

• receiving, from a user device via a graphical user interface, user information and subject tax return information;
• determining, by a processing element, a plurality of confidence indicators regarding the user,
• wherein each confidence indicator is indicative that the user is either genuine or fraudulent,
• wherein said step of determining a plurality of confidence indicators is performed by—
• acquiring a plurality of subject nodes related to the user information and the subject tax return information,
• comparing the plurality of subject nodes to a set of external nodes,
• linking at least one subject node to at least one fraud-indicating node through at least one intermediary node,
• wherein said fraud-indicating node and said intermediary node are within the set of external nodes;
• analyzing, by the processing element, said plurality of confidence indicators,
• determining, by the processing element, a likelihood that the user is genuine;
• and
• filing, with a computing device associated with a taxing authority, the subject tax return upon said likelihood being over a threshold.

A method of determining whether a subject tax return is fraudulent, the method comprising the following steps:

• generate, by a processing element, a plurality of external nodes based upon previously filed tax returns,
• wherein at least a portion of the external nodes are fraud-indicative nodes and at least a portion of the external nodes are neutral nodes;
• link, by the processing element, the plurality of external nodes together via shared relationships of related information;
• acquire, from a user device via a graphical user interface, tax information related to the subject tax return;
• identify, by the processing element, subject nodes based upon the tax information;
• compare, by the processing element, the subject nodes to external nodes for related information;
• link, by the processing element, at least one subject node to at least one fraud-indicative node via at least one neutral node;
• determine, based on said linking of at least one subject note to at least one fraud-indicative node, a likelihood that the user is fraudulent;
• and
• prevent filing, upon said likelihood being over a threshold, of the subject tax with a taxing authority.