Filtering Tweets for Social Unrest

Abstract

Abstract-Since the events of the Arab Spring, there has been increased interest in using social media to anticipate social unrest. While efforts have been made toward automated unrest prediction, we focus on filtering the vast volume of tweets to identify tweets relevant to unrest, which can be provided to downstream users for further analysis. We train a supervised classifier that is able to label Arabic language tweets as relevant to unrest with high reliability. We examine the relationship between training data size and performance and investigate ways to optimize the model building process while minimizing cost. We also explore how confidence thresholds can be set to achieve desired levels of performance.

Support Vector Machine Active Learning Algorithms with Query-by-Committee Versus Closest-to-Hyperplane Selection

“…If past information or relevant experience is available that enables an educated estimate of p, then that estimate may be used. If past information or relevant experience is not available, then p = 0.5 can be used since p = 0.5 will never underestimate n. This can be seen from Equation 6 where the product p(1 − p) occurs in the numerator and noting that the product p(1 − p) is maximized at p = 0.5.…”

“…Note that the first term on the right-hand side of Equation 9 is the estimate of the sample size that would be obtained using Equation 6, which assumed an infinite population and thus did not employ a finite population correction factor. Define n 0 to be this uncorrected sample size estimate, i.e.,…”

“…The use of active learning has received a lot of interest for reducing annotation costs for text and speech processing applications [1], [2], [3], [4], [5], [6]. Many applications have the following three characteristics: 1) they have imbalanced data sets, 2) training data annotation is a burden, and 3) support vector machines (SVMs) are able to train highperforming systems for the application.…”

The Use of Unlabeled Data Versus Labeled Data for Stopping Active Learning for Text Classification

“…Active learning has been shown to have a particularly large potential for reducing annotation cost for text classification [6], [7]. Text classification is one of the most important fields in semantic computing and it has been used in many applications [8], [9], [10], [11], [12].…”

Impact of Batch Size on Stopping Active Learning for Text Classification

“…The use of active learning has received a lot of interest for reducing annotation costs for text classification [1], [2], [3].…”