|dc.description.abstract||The security state of the country has deteriorated over time with an attribution to the low police to population ratio currently at 1:1150. Kenyan households have suffered the worst given the poor focus given to them with the assumption that they are accountable of their own security. This has caused many household owners to turn to the assistance of the many security agencies, most of which still employ outdated means of implementing security. These outdated means include physically stationing ill-equipped personnel in household vicinities to monitor and provide security response in case any arises. Their deficiency in equipment has, however, exposed households to the same initial dangers of insecurity.
In this study, I propose a Mobile-Based Security Agency Monitor and Alert System. This system relies on the integration of the vital solutions of Natural Language Processing (NLP) in voice and sound recognition, GPS location services, and message broadcasting to detect sound variations in the environment and alert security agents at the user end. An automatic analysis of the recorded sound is responsible for determining the need for a notification on the admin end. Abnormal sound variations based on the pitch measured in decibels on the user end will be responsible for alert notifications on the admin end that will trigger immediate response by security guards on the ground.
This innovation is aimed at eliminating the need of physically stationing security agents at households, thus reducing any excesses in expenditure, and implement a strategy to only deploy response when needed. This will improve effectiveness and reduce cases of inadequacy of security guards. The innovation also facilitates an alert broadcast system in case of security breaches that does not require user’s manual manipulation or prompting, motivated by the impulse reaction by humans in drastic situations to produce high-pitched voices.
The testing of the implemented application took place among the respondents that were sampled in earlier stages of user requirement analysis. After initial education, distribution, and installation, testers were informed to focus on issues of functionality, user interface, learnability, and usability. Sampling of the eventual testing results indicated a positive response to the implementation, displaying a good acceptability margin, and potential of quick adoption among users.||en_US