System for Identification of Voice Calls of Interest in a Telecom Communication Network

In recent times, there has been a huge increase in the number of value-added services being provided to end customers. This has been further complemented by the advancement in telecommunication services including the introduction of 5G and associated services provided by communication service providers (CSPs) or Over the Top (OTT) applications vendors. As a result of this increase in services, different targeted marketing campaigns are used by various vendors to approach the end customers through different communications channels e.g., voice, data, SMS, USSD, etc. Specifically, amongst these, marketing through voice channels is important considering it to be a more real-time communication where the end-users need to act (accept or reject the call) in a specific time to establish the communication, unlike SMS wherein response can be more asynchronous. Considering the above, it becomes even more important for the end-users to identify if the received voice call is of interest to the user or is a Spam call. This has been a long-standing problem to accurately distinguish between Spam calls and ‘Call of Interest’. ‘Call of Interest’ may refer to a voice call that the end-user would not like to miss but it may be from someone not available in saved contacts (e.g., a response call from a repair agency that was contacted earlier by the end-user). Various methods are available in the state of the art to solve the said problem; however, those may not be sufficient to accurately identify the ‘Call of Interest’. As mentioned, there can be instances wherein the call may be a Spam call for a set of users however it may be of interest to another user. Accurate identification of such a ‘Call of Interest’ is the focus of this paper.