Context aware ontology-based hybrid intelligent framework for vehicle driver categorization

In public vehicles, one of the major concerns is driver's level of expertise for its direct proportionality to safety of passengers. Hence, before a driver is subjected to certain type of vehicle, he should be thoroughly evaluated and categorized with respect to certain parameters instead of only one-time metric of having driving license. These aspects may be driver's expertise, vigilance, aptitude, experience years, cognition, driving style, formal education, terrain, region, minor violations, major accidents, and age group. The purpose of this categorization is to ascertain suitability of a driver for certain vehicle type(s) to ensure passengers' safety. Currently, no driver categorization technique fully comprehends the implicit as well as explicit characteristics of drivers dynamically. In this paper, machine learning-based dynamic and adaptive technique named D-CHAITs (driver categorization through hybrid of artificial intelligence techniques) is proposed for driver categorization with an objective focus on driver's attributes modeled in DriverOntology. A supervised mode of learning has been employed on a labeled dataset, having diverse profiles of drivers with attributes pertinent to drivers' perspectives of demographics, behaviors, expertise, and inclinations. A comparative analysis of D-CHAIT with three other machine learning techniques (fuzzy logic, case-based reasoning, and artificial neural networks) is also presented. The efficacy of all techniques was empirically measured while categorizing the drivers based on their profiles through metrics of accuracy, precision, recall, F-measure performance, and associated costs. These empirical quantifications assert D-CHAIT as a better technique than contemporary ones. The novelty of proposed technique is signified through preprocessing of feature attributes, quality of data, training of machine learning model on more relevant data, and adaptivity.

Automated summary

This study proposes a dynamic and adaptive machine learning technique called D-CHAITs for driver categorization. By modeling driver attributes, this technique aims to categorize drivers based on their profiles. Comparative analysis demonstrates that D-CHAIT outperforms other techniques. The novelty of this technique lies in its preprocessing of feature attributes, data quality, training on relevant data, and adaptivity.