-
Echo chamber formation sharpened by priority users
Authors:
Henrique F. de Arruda,
Kleber A. Oliveira,
Yamir Moreno
Abstract:
Priority users (e.g., verified profiles on Twitter) are social media users whose content is promoted by recommendation algorithms. However, the impact of this heterogeneous user influence on opinion dynamics, such as polarization phenomena, is unknown. We conduct a computational mechanistic investigation of such consequences in a stylized setting. First, we allow priority users, whose content has…
▽ More
Priority users (e.g., verified profiles on Twitter) are social media users whose content is promoted by recommendation algorithms. However, the impact of this heterogeneous user influence on opinion dynamics, such as polarization phenomena, is unknown. We conduct a computational mechanistic investigation of such consequences in a stylized setting. First, we allow priority users, whose content has greater reach (similar to algorithmic boosting), into an opinion model on adaptive networks. Then, to exploit this gain in influence, we incorporate stubborn user behavior, i.e., zealot users who remain committed to opinions throughout the dynamics. Using a novel measure of echo chamber formation, we find that prioritizing users can inadvertently reduce polarization if they post according to the same rule but sharpen echo chamber formation if they behave heterogeneously. Moreover, we show that a minority of extremist ideologues (i.e., users who are both stubborn and priority) can push the system into a transition from consensus to polarization with echo chambers. Our findings imply that the implementation of the platform's prioritization policy should be carefully monitored in order to ensure there is no abuse of users with extra influence.
△ Less
Submitted 14 December, 2023;
originally announced December 2023.
-
An Evidence-based Roadmap for IoT Software Systems Engineering
Authors:
Rebeca C. Motta,
Káthia M. de Oliveira,
Guilherme H. Travassos
Abstract:
Context: The Internet of Things (IoT) has brought expectations for software inclusion in everyday objects. However, it has challenges and requires multidisciplinary technical knowledge involving different areas that should be combined to enable IoT software systems engineering. Goal: To present an evidence-based roadmap for IoT development to support developers in specifying, designing, and implem…
▽ More
Context: The Internet of Things (IoT) has brought expectations for software inclusion in everyday objects. However, it has challenges and requires multidisciplinary technical knowledge involving different areas that should be combined to enable IoT software systems engineering. Goal: To present an evidence-based roadmap for IoT development to support developers in specifying, designing, and implementing IoT systems. Method: An iterative approach based on experimental studies to acquire evidence to define the IoT Roadmap. Next, the Systems Engineering Body of Knowledge life cycle was used to organize the roadmap and set temporal dimensions for IoT software systems engineering. Results: The studies revealed seven IoT Facets influencing IoT development. The IoT Roadmap comprises 117 items organized into 29 categories representing different concerns for each Facet. In addition, an experimental study was conducted observing a real case of a healthcare IoT project, indicating the roadmap applicability. Conclusions: The IoT Roadmap can be a feasible instrument to assist IoT software systems engineering because it can (a) support researchers and practitioners in understanding and characterizing the IoT and (b) provide a checklist to identify the applicable recommendations for engineering IoT software systems.
△ Less
Submitted 14 March, 2023;
originally announced March 2023.
-
Text-Driven Video Acceleration: A Weakly-Supervised Reinforcement Learning Method
Authors:
Washington Ramos,
Michel Silva,
Edson Araujo,
Victor Moura,
Keller Oliveira,
Leandro Soriano Marcolino,
Erickson R. Nascimento
Abstract:
The growth of videos in our digital age and the users' limited time raise the demand for processing untrimmed videos to produce shorter versions conveying the same information. Despite the remarkable progress that summarization methods have made, most of them can only select a few frames or skims, creating visual gaps and breaking the video context. This paper presents a novel weakly-supervised me…
▽ More
The growth of videos in our digital age and the users' limited time raise the demand for processing untrimmed videos to produce shorter versions conveying the same information. Despite the remarkable progress that summarization methods have made, most of them can only select a few frames or skims, creating visual gaps and breaking the video context. This paper presents a novel weakly-supervised methodology based on a reinforcement learning formulation to accelerate instructional videos using text. A novel joint reward function guides our agent to select which frames to remove and reduce the input video to a target length without creating gaps in the final video. We also propose the Extended Visually-guided Document Attention Network (VDAN+), which can generate a highly discriminative embedding space to represent both textual and visual data. Our experiments show that our method achieves the best performance in Precision, Recall, and F1 Score against the baselines while effectively controlling the video's output length. Visit https://www.verlab.dcc.ufmg.br/semantic-hyperlapse/tpami2022/ for code and extra results.
△ Less
Submitted 29 March, 2022;
originally announced March 2022.
-
An influencer-based approach to understanding radical right viral tweets
Authors:
Laila Sprejer,
Helen Margetts,
Kleber Oliveira,
David O'Sullivan,
Bertie Vidgen
Abstract:
Radical right influencers routinely use social media to spread highly divisive, disruptive and anti-democratic messages. Assessing and countering the challenge that such content poses is crucial for ensuring that online spaces remain open, safe and accessible. Previous work has paid little attention to understanding factors associated with radical right content that goes viral. We investigate this…
▽ More
Radical right influencers routinely use social media to spread highly divisive, disruptive and anti-democratic messages. Assessing and countering the challenge that such content poses is crucial for ensuring that online spaces remain open, safe and accessible. Previous work has paid little attention to understanding factors associated with radical right content that goes viral. We investigate this issue with a new dataset ROT which provides insight into the content, engagement and followership of a set of 35 radical right influencers. It includes over 50,000 original entries and over 40 million retweets, quotes, replies and mentions. We use a multilevel model to measure engagement with tweets, which are nested in each influencer. We show that it is crucial to account for the influencer-level structure, and find evidence of the importance of both influencer- and content-level factors, including the number of followers each influencer has, the type of content (original posts, quotes and replies), the length and toxicity of content, and whether influencers request retweets. We make ROT available for other researchers to use.
△ Less
Submitted 15 September, 2021;
originally announced September 2021.
-
IoT Roadmap: Support for Internet of Things Software Systems Engineering
Authors:
Rebeca Motta,
Káthia Oliveira,
Guilherme Travassos
Abstract:
The Roadmap is performed in the context of a Ph.D. research in collaboration between the Experimental Software Engineering Group, from the Systems Engineering and Computing Program of the Federal University of Rio de Janeiro (COPPE/UFRJ) and the Laboratory of Industrial and Human Automation Control, Mechanical engineering and Computer Science (LAMIH UMR CNRS 8201) in the Universitè Polytechnique H…
▽ More
The Roadmap is performed in the context of a Ph.D. research in collaboration between the Experimental Software Engineering Group, from the Systems Engineering and Computing Program of the Federal University of Rio de Janeiro (COPPE/UFRJ) and the Laboratory of Industrial and Human Automation Control, Mechanical engineering and Computer Science (LAMIH UMR CNRS 8201) in the Universitè Polytechnique Hauts-de-France (UPHF).
The Roadmap resulted from an investigation on the particularities of IoT applications. It is the concrete organization of the concepts and evidence gathered from different experimental studies. It comes to support the definition of IoT software systems, with specific items for the project team to discuss and define the essential aspects related to the specifying, designing, and implementing an IoT application.
△ Less
Submitted 12 July, 2021; v1 submitted 8 March, 2021;
originally announced March 2021.
-
Technical Report: Rapid Reviews on Engineering of Internet of Things Software Systems
Authors:
Rebeca Motta,
Káthia de Oliveira,
Guilherme Travassos
Abstract:
We conducted a set of Rapid Reviews to characterize Internet of Things facets. We formatted a generic meta-protocol that was instantiated for each of the six facets presented (Connectivity, Things, Behavior, Smartness, Interactivity, and Environment)and considering the issue of Security, one of the most important and frequent challenges in the context of IoT. The meta-protocol is detailed and the…
▽ More
We conducted a set of Rapid Reviews to characterize Internet of Things facets. We formatted a generic meta-protocol that was instantiated for each of the six facets presented (Connectivity, Things, Behavior, Smartness, Interactivity, and Environment)and considering the issue of Security, one of the most important and frequent challenges in the context of IoT. The meta-protocol is detailed and the results of each review are presented.
△ Less
Submitted 14 January, 2021;
originally announced January 2021.
-
LRCN-RetailNet: A recurrent neural network architecture for accurate people counting
Authors:
Lucas Massa,
Adriano Barbosa,
Krerley Oliveira,
Thales Vieira
Abstract:
Measuring and analyzing the flow of customers in retail stores is essential for a retailer to better comprehend customers' behavior and support decision-making. Nevertheless, not much attention has been given to the development of novel technologies for automatic people counting. We introduce LRCN-RetailNet: a recurrent neural network architecture capable of learning a non-linear regression model…
▽ More
Measuring and analyzing the flow of customers in retail stores is essential for a retailer to better comprehend customers' behavior and support decision-making. Nevertheless, not much attention has been given to the development of novel technologies for automatic people counting. We introduce LRCN-RetailNet: a recurrent neural network architecture capable of learning a non-linear regression model and accurately predicting the people count from videos captured by low-cost surveillance cameras. The input video format follows the recently proposed RGBP image format, which is comprised of color and people (foreground) information. Our architecture is capable of considering two relevant aspects: spatial features extracted through convolutional layers from the RGBP images; and the temporal coherence of the problem, which is exploited by recurrent layers. We show that, through a supervised learning approach, the trained models are capable of predicting the people count with high accuracy. Additionally, we present and demonstrate that a straightforward modification of the methodology is effective to exclude salespeople from the people count. Comprehensive experiments were conducted to validate, evaluate and compare the proposed architecture. Results corroborated that LRCN-RetailNet remarkably outperforms both the previous RetailNet architecture, which was limited to evaluating a single image per iteration; and a state-of-the-art neural network for object detection. Finally, computational performance experiments confirmed that the entire methodology is effective to estimate people count in real-time.
△ Less
Submitted 12 May, 2020; v1 submitted 20 April, 2020;
originally announced April 2020.
