New York
LondonBeijing
Intelligent Speech Processing Technique for Suspicious Voice Call Identification Using Adaptive Machine Learning Approach

Intelligent Speech Processing Technique for Suspicious Voice Call Identification Using Adaptive Machine Learning Approach

Dhilip Kumar (Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, India), Swathi P. (Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, India), Ayesha Jahangir (Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, India), Nitesh Kumar Sah (Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology, India), and Vinothkumar V. (MVJ College of Engineering, India)
Copyright: © 2021 |Pages: 9
DOI: 10.4018/978-1-7998-6870-5.ch025
OnDemand:
(Individual Chapters)
Available
$37.50
Current Special Offers
No Current Special Offers
TOTAL SAVINGS: $37.50

Abstract

With recent advances in the field of data, there are many advantages of speedy growth of internet and mobile phones in the society, and people are taking full advantage of them. On the other hand, there are a lot of fraudulent happenings everyday by stealing the personal information/credentials through spam calls. Unknowingly, we provide such confidential information to the untrusted callers. Existing applications for detecting such calls give alert as spam to all the unsaved numbers. But all calls might not be spam. To detect and identify such spam calls and telecommunication frauds, the authors developed the application for suspicious call identification using intelligent speech processing. When an incoming call is answered, the application will dynamically analyze the contents of the call in order to identify frauds. This system alerts such suspicious calls to the user by detecting the keywords from the speech by comparing the words from the pre-defined data set provided to the software by using intelligent algorithms and natural language processing.
Chapter Preview
Top

1. Introduction

With the development of the growing technology day by day, people are enjoying various sorts of services from the web. There are nearly 4 billion Internet users in 2018 (nearly half of the world’s population of 7.7 billion), up from 2 billion in 2015. Their private information is gradually leaked out. Like street crime, which historically grew in reference to increase, we are witnessing an identical evolution of cyber-crime. If a person’s privacy is held by attackers, he might be the target of telecommunication frauds. Latest statistics in 2019 show that 90 In order to detect suspicious calls, most of the current approaches are supported on labeling the caller numbers that are identified as spam by customers. There are also many researchers who use machine learning and other techniques to detect fraud calls. The main objective of the proposed work is to alert the mobile users to prevent from spam calls. This methodology helps to protect their credential information from cyber criminals.

1.1 Scope

These days most of the online banking users, unknowingly provide their credentials to the untrusted parties. The uneducated and unaware people usually face these kinds of problems. To overcome this, our proposed speech processing techniques helps to give alert / Risk notification to the users when untrusted parties talk about the credentials. So that we can avoid those kinds of SPAM calls through the machine learning techniques. Even if the user is unaware of the call being fraud, the app automatically detects it and ask for disconnection of call with waning.

1.2 Limitations

There are some limitations to our project like the voice should be loud and clear. The phone should have google speech to text API which is pre-installed in all google supported phones. If the sentence has minimum of 2 suspicious words from the dataset then only the user is alerted of the suspicious call. Though it does not require Internet as it uses google api but there should be at least good network for the voice to be clear. Multilingual languages and vocabulary of word is one of the critical issues to identify the suspicious word and it also needs large number of datasets to train the HMM model.

Complete Chapter List

Search this Book:
Reset