Big data learning and suggestions in modern apps

14th International Conference on Science, Engineering and Technology

自然科学;工业技术

Sharma, G.^1 ; Nadesh, R.K.^1 ; Arivuselvan, K.^1

School of Information Technology and Engineering, VIT University, Vellore

Tamil Nadu

632014, India^1

关键词: Apache storms;    Location-based applications;    Machine learning models;    MongoDB;    Prediction-based;    Probabilistic modeling;    Real time;    Realtime processing;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/263/4/042074/pdf
DOI  :  10.1088/1757-899X/263/4/042074

来源: IOP

【 摘 要 】

Among many other tasks involved for emergent location-based applications such as those involved in prescribing touring places and those focused on publicizing based on destination, destination prediction is vital. Dealing with destination prediction involves determining the probability of a location (destination) depending on historical trajectories. In this paper, a destination prediction based on probabilistic model (Machine Learning Model) feed-forward neural networks will be presented, which will work by making the observation of driver's habits. Some individuals drive to same locations such as work involving same route every day of the working week. Here, streaming of real-time driving data will be sent through Kafka queue in apache storm for real-time processing and finally storing the data in MongoDB.