Social media have been increasingly used by different users to explore their opinions on particular or social related concept in short text message communication. Due to the widespread utilization of social related networks to exchange and transform their opinions and data.A general concept relates to public, media relates to news and news relates to politics. It has been enabled a new research in political science. Predicting political lenience of unknown data elements both explicit and implicit, influence of opinions relates to political news. In this phenomena, predict the popularity of particular person by exploring varied twitter data.The expressions of different user’s opinions, predict performance and behavior of political actor by categorizing them as either positive or negative. Tensor flow based sentiment classifier is used to explore the data related to public view in terms of negative and positive. This can be used to predict the political member popularity. Because of data complexity, high sentiment influence of different active users which are involved in sharing data and more time complexity for multi – labeled instances assigned to multiple categories. Active learning concept is introduced to handle labeling effort in single attribute categorization. In this paper, propose a Novel Active Learning based Multi Labeled Categorization (NALMLC) strategy to support adaptive multi-labeled active learning. Proposed approach consist two stages, in first stage, it perform one-to many relation between different objects or same objects. In second stage, use mule class label procedure to obtain meaningful and non-relative example. Proposed approach finds the relations between different entities using iterative processor, for that it requires different entities for single regularization. Experimental results show that proposed approach performs well on different scenarios, accurate instance association with single regularization. This increases the complexity while describing the results.
Volume 12 | Issue 6
Pages: 1467-1476
DOI: 10.5373/JARDCS/V12I2/S20201344