Digital Platform for Local Service and Connection System using Deep learning and Multilingual Voice Search.

This paper has presented a structured and critical literature review of six studies relevant to a proposed digital platform for local service and connection. The platform concept, which remains at the study and review phase, intends to integrate CNN based face verification, machine learning job recommendation, and multilingual voice search into a single cohesive system. No implementation has been conducted, no datasets have been trained, and no experimental results are reported. The reviewed studies collectively demonstrate that deep learning methods are well suited to the face verification task; that hybrid recommendation systems offer a practical path around the cold-start problem; and that multimodal NLP provides the methodological foundation for voice-based interaction. Ranjan et al. [1] and Thai et al. [4] contribute directly to the verification component; Mitra et al. [2] provide essential security and privacy guidance; Wei [3] offers a video-based enhancement that may be incorporated at a later stage; Ali et al. [6] provide the recommendation design framework and ethics guidance; and Shende et al. [5] inform the NLP and multimodal interaction design. The comparative analysis identified four key gaps in the reviewed literature: the absence of integrated multi-component architectures for local service platforms; inadequate evaluation of face verification under real-world mobile conditions; underexplored recommendation methods for informal labour markets; and insufficient attention to low-resource and regional language voice interaction. These gaps define the specific contributions that the proposed system’s implementation phase will aim to address. The proposed framework intends to fill these gaps by developing an integrated platform whose three components are informed by the reviewed evidence, whose architecture reflects the practical constraints of the target deployment environment, and whose design gives explicit attention to the fairness, privacy, and accessibility requirements that the reviewed literature collectively highlights as important but insufficiently resolved. Future implementation results will be reported in subsequent work.

[1]          R. Ranjan, S. Sankaranarayanan, A. Bansal, N. Bodla, J.-C. Chen, V. M. Patel, C. D. Castillo, and R. Chellappa, “Deep learning for understanding faces: machines may be just as good, or better, than humans,” IEEE Signal Processing Magazine, vol. 35, no. 1, pp. 66–83, Jan. 2018. [2]          A. Mitra, S. Gupta, and P. Mukherjee, “iFace 1.1: a facial authentication based digital identity system for smart city applications using cloud integrated biometric verification,” in Proc. IEEE Intl. Conf. on Smart Systems and Inventive Technology (ICSSIT), pp. 1–7, 2021. [3]          J. Wei, “Deep learning-based video face recognition for secure digital transaction platforms using MTCNN and convolutional feature extraction,” in Proc. IEEE Intl. Conf. on Computer Vision and Machine Intelligence (CVMI), pp. 1–6, 2020. [4]          H.-D. Thai, T.-N. Nguyen, and V.-H. Le, “A low-cost facial recognition mobile application for employee attendance monitoring using ResNet34 and cloud-based architecture,” in Proc. Intl. Conf. on Advanced Technologies for Communications (ATC), pp. 1–6, 2021. [5]          R. H. Shende and A. A. Shinde, “A review of machine learning and deep learning techniques for sign language recognition using CNNs, image processing, speech recognition, and natural language processing,” International Journal of Advanced Computer Science and Applications (IJACSA), vol. 13, no. 4, pp. 1–10, 2022. [6]          M. Ali, S. Iqbal, and A. Khan, “Artificial intelligence applications in intelligent digital systems: a review of machine learning, deep learning, and NLP for recommendation, personalisation, and ethical AI,” Journal of Intelligent Systems and Internet of Things (JISIoT), vol. 7, no. 2, pp. 45–62, 2022.