Verizon Communications Inc.
Chat analysis using machine learning

Abstract:

A device may receive information associated with a set of chat logs. The device may obtain context information associated with the information, wherein the context information identifies a network address associated with a participant of the set of chat logs. The device may determine whether the set of chat logs is to be assigned to a first category, a second category, or a third category, wherein the first category is associated with fraudulent chat logs, wherein the second category is associated with chat logs involving a misrepresentation, and wherein the third category is associated with chat logs that are not identified as fraudulent or involving a misrepresentation. The device may perform an action based on whether the set of chat logs is assigned to the first category, the second category, or the third category.