Prevention of Unwanted Calls Over Telephony Network
Gaurav Malviya1, Vatsal Singh2

1Gaurav Malviya, UG Student Computer Science and Engineering SRM Institute of Science and Technology, Chennai, India.
2Vatsal Singh, UG Student Computer Science and Engineering SRM Institute of Science and Technology, Chennai, India.

Manuscript received on May 17, 2020. | Revised Manuscript received on May 22, 2020. | Manuscript published on May 30, 2020. | PP: 2592-2595 | Volume-9 Issue-1, May 2020. | Retrieval Number: F8972038620/2020©BEIESP | DOI: 10.35940/ijrte.F8972.059120
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: The main motive is to increase the malicious call prevention while not relying on any underlying architecture of the phone . Challenges faced here are -how to gather useful information on how to decrease benign calls from being blocked . The first part of the work is to gather data concerning nuisance callers so as to create an efficient interference mechanism, with the help of Machine Learning Algorithms. Gathering of the dataset is based on client statistics. In that, there are three varieties of lessons like telemarketing, Unwanted (Malicious), robocalls.If a call is Malicious or not is judged based on few functions .So compare the effectiveness of these capabilities with the aid of numerous cutting-edge models. For this, usage of both Area under curve (AUC) score, and layout a brand new metric, average first prediction (AFP) is done. Average first prediction is designed to assess the averaged amount of nuisance calls that desires to be located before a user can expect it as a nuisance caller, without affecting benign traffic. Evaluation shows that segregating the planned options, a random forest version can do Associate in Nursing AUC score of at 0.98 ; additionally , it reduces the averaged important determined nuisance calls via up to 88% from a black-listing method, whilst making sure that over 98% of the benign calls will now no longer be banned from connecting . The system would be fast, effective, efficient, UI would be lightweight ,and this would work without accessing the telephony network architecture. 
Keywords: Robocalls, telephony network, unwanted calls, network calls, voice-over-IP.
Scope of the Article: Network Architecture