Peace processes are complex, protracted, and contentious involving significant bargaining and compromising among various societal and political stakeholders. In civil war terminations, it is pertinent to measure the pulse of the nation to ensure that the peace process is responsive to citizens' concerns. Social media yields tremendous power as a tool for dialogue, debate, organization, and mobilization thereby adding more complexity to the peace process. Using Colombia's final peace agreement and national referendum as a case study, we investigate the influence of two important indicators: inter-group polarization and public sentiment towards the peace process. We present a detailed linguistic analysis to detect inter-group polarization and a predictive model that leverages tweet structure, content, and user-based features to predict public sentiment towards the Colombian peace process. We demonstrate that had pro-accord stakeholders leveraged public opinion from social media the outcome of the Colombian referendum could have been different.
Peace processes are complex, protracted, and contentious involving significant bargaining and compromising among various societal and political stakeholders. In civil war terminations, it is pertinent to measure the pulse of the nation to ensure that the peace process is responsive to citizens' concerns. Social media yields tremendous power as a tool for dialogue, debate, organization, and mobilization thereby adding more complexity to the peace process. Using Colombia's final peace agreement and national referendum as a case study, we investigate the influence of two important indicators: inter-group polarization and public sentiment towards the peace process. We present a detailed linguistic analysis to detect inter-group polarization and a predictive model that leverages tweet structure, content, and user-based features to predict public sentiment towards the Colombian peace process. We demonstrate that had pro-accord stakeholders leveraged public opinion from social media the outcome of the Colombian referendum could have been different.
Twitter provides a platform for information sharing and diffusion, and has quickly emerged as a mechanism for organizations to engage with their consumers. A driving factor for engagement is providing relevant and timely content to users. We posit that the engagement via tweets offers a good potential to discover user interests and leverage that information to target specific content of interest. To that end, we have developed a framework that analyzes tweets to identify the interests of current followers and leverages topic models to deliver a personalized topic profile for each user. We validated our framework by partnering up with a local media company and analyzing the content gap between them and their followers. We also developed a mobile application that incorporates the proposed framework.
There is an increasing trend amongst users to consume information from websites and social media. With the huge influx of content it becomes challenging for the consumers to navigate to topics or articles that interest them. Particularly in health care, the content consumed by a user is controlled by various factors such as demographics and lifestyle. In this paper, we use a semi-bipartite network model to capture the interactions between users and health topics that interest them. We use a supervised link prediction approach to recommend topics to users based on their past reading behavior and contextual data associated to a user such as demographics.
As robots enter human spaces, unique perception challenges are emerging. Sensing human activity, adapting to highly dynamic environments, and acting coherently and contingently is challenging when robots transition from structured environments to human-centric ones. We approach this problem by employing context-based perception, a biologically-inspired, low-cost approach to sensing that leverages noisy, global features. Across several months, our mobile robot collected real-world, multimodal data from multi-use locations; where the same space might be used for many different activities. We then ran a series of unimodal and multimodal classification experiments. We successfully classified several aspects of situational context from noisy data, and, to our knowledge are the first group to do so. This work represents an important step toward enabling robots that can readily leverage context to solve perceptual tasks.
Face verification, though for humans seems to be an easy task, is a long-standing research area. With challenging covariates such as disguise or face obfuscation, automatically verifying the identity of a person is assumed to be very hard. This paper explores the feasibility of face verification under disguise variations using multi-spectrum (visible and thermal) face images. We propose a framework, termed as Anavrta, which classifies the local facial regions of both visible and thermal face images into biometric (regions without disguise) and non-biometric (regions with disguise) classes. The biometric patches are then used for facial feature extraction and matching. The performance of the algorithm is evaluated on the IIITD In and Beyond Visible Spectrum Disguise database that is prepared by the authors and contains images pertaining to 75 subjects with different kinds of disguise variations. The experimental results suggest that the proposed framework improves the performance compared to existing algorithms, however there is a need for more research to address this important covariate.
Stochastic Gradient Descent (SGD) is a simple yet very efficient online learning algorithm for optimizing convex (and often non-convex) functions and one of the most popular stochastic optimization methods in machine learning today. One drawback of SGD is that it is sensitive to the learning rate hyper-parameter. The Adaptive Sub-gradient Descent, AdaGrad, dynamically incorporates knowledge of the geometry of the data observed in earlier iterations to calculate a different learning rate for every feature. In this work, we implement a distributed version of AdaGrad for large-scale machine learning tasks using Apache Spark. Apache Spark is a fast cluster computing engine that provides similar scalability and fault tolerance properties to MapReduce, but in contrast to Hadoop's two-stage disk-based MapReduce paradigm, Spark's multi-stage in-memory primitives allow user programs to load data into a cluster's memory and query it repeatedly, which makes it ideal for building scalable machine learning applications. We empirically evaluate our implementation on large-scale real-world problems in the machine learning canonical tasks of classification and regression. Comparing our implementation of AdaGrad with the SGD scheduler currently available in Spark's Machine Learning Library (MLlib), we experimentally show that AdaGrad saves time by avoiding manually setting a learning-rate hyperparameter, converges fast and can even achieve better generalization errors.
Twitter is a micro-blogging site which enables individuals to write about their daily activities, express opinions, share information and connect with other users and businesses. Twitter is widely used by companies or brands to reach out to their customers, increase awareness about various topics and products. According to a Twitter report1, there are 284 million monthly active users, and 500 million tweets shared every day. This provides a unique business opportunity for companies to reach out to their customers, increase awareness and interaction among users about a topic or product and keep them engaged with their content. Using the 80/20 principle2, 80% of the Tweets should be directed towards interacting and engaging with the current followers. The relevant information cannot be captured using only the hashtags, mentions or keyword search, we need to model user’s interests based on their tweeting patterns and the content of the Tweet. The companies need to listen to the trending topics amongst their followers to connect with them. In order to discover content a user is or might be interested in, we need to build a topic profile for each user which contains high level topics which he/she discusses in the post3.