Deep learning based next word prediction aided assistive gaming technology for people with limited vocabulary

Abstract

AimVocabulary is the base for communication and vocabulary deficits reduce effective social communication. This study aims to develop and analyze English language-based assistive gaming technology to support individuals with vocabulary deficits. Materials and methodsThe developed intervention is based on deep learning technology for next word prediction, and it is equipped with web and mobile application interaction support. Various deep learning models, including long short-term memory, bidirectional long short-term memory, and bidirectional encoder representations from transformers, were learned and deployed using an open-source dataset corpus. The proposed intervention was analyzed using a specified task of a 5-minute lecture on a simple technical topic. The completed activities were also evaluated by professionals using vocabulary-related expert characteristics. ResultsThe results of the conducted 5-minutetask on social domain without intervention and later with intervention improved. The results were assessed by counting vocabulary words as well as by vocabulary specialists. The levels of centraltendencies and deviations changed as the task improved with the proposed intervention. Mean, median, and standard deviation scores of the experimental group improved significantly (189.1, 194.0, and 67.61) in comparison to the control group (125.4, 127.0, and 43.55). Furthermore, language specialists observed significant differences in the mean and median without intervention (12.84 and 13.0) compared to the intervention group (24.23 and 24.0, respectively). ConclusionsThe outcome of the experiments appears to be in favor of the suggested intervention. As a result, the suggested deep learning based next word prediction aided assistive gaming technology could support persons in overcoming their limited vocabulary disadvantages. In the future, the integration of the proposed technology into existing communication enhancement technology could achieve better results.