Deep Learning Model Development to Classify Technology from Social Media

Abstract

The important characteristic of societal platform is the unremitting cohort of content which leads to root of novel data out of it. Community platforms are active in nature and they can be painstaking as new type of data resources for upcoming trend forecasts with application of data analytics technique. Due to world-wide contribution in open-source technologies, new technologies are released frequently. The innovation and instruments learned by individuals become supplanted in a brief timeframe. As the I.T. industry required regular overhauls in abilities and new advancements are being discharged, it is basic to trail and realize novel upcoming innovation patterns of the field. To accomplish this point, a deep learning model is created to distinguish up and coming innovations from internet-based life strings. This paper presents technology classifier model to categorize unstructured text content in to relevant technologies. To develop technology classifier, classification algorithms SVM - Support Vector Machine, Decision Tree, kNN – k Nearest Neighbor, Artificial Neural Network and Deep Feed-Forward Neural Network are trained and experimented to forecast expertise terms from unstructured content of social posts. The results of experiment show that Feed-Forward deep neural network outperforms other classification models and provide better accuracy and technology term prediction.