1,721,027 research outputs found
Profiling Cryptocurrency Influencers with Few-Shot Learning Using Data Augmentation and ELECTRA
Full text linkWith this work we propose an application of the ELECTRA Transformer, fine-tuned on two augmented version of the same training dataset. Our team developed the novel framework for taking part at the Profiling Cryptocurrency Influencers with Few-shot Learning task hosted at PAN@CLEF2023. Our proposed strategy consists of an early data augmentation stage followed by a fine-tuning of ELECTRA. At the first stage we augment the original training dataset provided by the organizers using backtranslation. Using this augmented version of the training dataset, we perform a fine tuning of ELECTRA. Finally, using the fine-tuned version of ELECTRA, we inference the labels of the samples provided in the test set. To develop and test our model we used a two-ways validation on the training set. Firstly, we evaluate all the metrics on the augmented training set, and then we evaluate on the original training set. The metrics we considered span from accuracy to Macro F1, to Micro F1, to Recall and Precision. According to the official evaluator, our best submission reached a Macro F1 value equal to 0.3762
Text Enrichment with Japanese Language to Profile Cryptocurrency Influencers
Full text linkFrom a few-shot learning perspective, we propose a strategy to enrich the latent semantic of the text provided in the dataset provided for the Profiling Cryptocurrency Influencers with Few-shot Learning, the task hosted at PAN@CLEF2023. Our approach is based on data augmentation using the backtranslation forth and back to and from Japanese language. We translate samples in the original training dataset to a target language (i.e. Japanese). Then we translate it back to English. The original sample and the backtranslated one are then merged. Then we fine-tuned two state-of-the-art Transformer models on this augmented version of the training dataset. We evaluate the performance of the two fine-tuned models using the Macro and Micro F1 accordingly to the official metric used for the task. After the fine-tuning phase, ELECTRA and XLNet obtained a Macro F1 of 0.7694 and 0.7872 respectively on the original training set. Our best submission obtained a Macro F1 equal to 0.3851 on the official test set provided
Semantically-aware statistical metrics via weighting kernels
No full textDistance metrics between statistical distributions are widely used as an efficient mean to aggregate/simplify the underlying probabilities, thus enabling high-level analyses. In this paper we investigate the collisions that can arise with such metrics, and a mitigation technique rooted on kernels. In detail, we first show that the existence of colliding functions (so-called iso-curves) is widespread across metrics and families of functions (e.g., gaussians, heavy-tailed). Later, we propose a solution based on kernels for augmenting distance metrics and summary statistics, thus avoiding collisions and highlighting semantically-relevant phenomena. This study is supported by a thorough theoretical evaluation of our solution against a large number of functions and metrics, complemented by a real-world evaluation carried out by applying our solution to an existing problem. Some further research venues are also discussed. The theoretical construction and the achieved results show the soundness, viability, and quality of our proposal that, other being interesting on its own, also paves the way for further research in the highlighted directions
MulBot: Unsupervised Bot Detection Based on Multivariate Time Series
No full textOnline social networks are actively involved in removing malicious social bots due to their role in spreading low-quality information. However, most of the existing bot detectors are supervised classifiers incapable of capturing the evolving behavior of sophisticated bots. Here we propose MulBot, an unsupervised bot detector based on multivariate time series (MTS). For the first time, we exploit multidimensional temporal features extracted from user timelines. We manage the multidimensionality with an LSTM autoencoder, which projects the MTS in a suitable latent space. Then, we perform a clustering step on this encoded representation to identify dense groups of very similar users - a known sign of automation. Finally, we perform a binary classification task achieving f1-score =0.99, outperforming state-of-the-art methods (f1-score ≤ 0.97). Not only does MulBot achieve excellent results in the binary classification task, but we also demonstrate its strengths in a novel and practically-relevant task: detecting and separating different botnets. In this multiclass classification task we achieve f1-score =0.96. We conclude by estimating the importance of the different features used in our model and by evaluating MulBot's capability to generalize to new unseen bots, thus proposing a solution to the generalization deficiencies of supervised bot detectors
$FAKE: Evidence of spam and bot activity in stock microblogs on twitter
No full textMicroblogs are increasingly exploited for predicting prices and traded volumes of stocks in financial markets. However, it has been demonstrated that much of the content shared in microblogging platforms is created and publicized by bots and spammers. Yet, the presence (or lack thereof) and the impact of fake stock microblogs has never systematically been investigated before. Here, we study 9M tweets related to stocks of the 5 main financial markets in the US. By comparing tweets with financial data from Google Finance, we highlight important characteristics of Twitter stock microblogs. More importantly, we uncover a malicious practice perpetrated by coordinated groups of bots and likely aimed at promoting low-value stocks by exploiting the popularity of high-value ones. Our results call for the adoption of spam and bot detection techniques in all studies and applications that exploit user-generated content for predicting the stock market
TweepFake: About detecting deepfake tweets
No full textThe recent advances in language modeling significantly improved the generative capabilities of deep neural models: In 2019 OpenAI released GPT-2, a pre-trained language model that can autonomously generate coherent, non-trivial and human-like text samples. Since then, ever more powerful text generative models have been developed. Adversaries can exploit these tremendous generative capabilities to enhance social bots that will have the ability to write plausible deepfake messages, hoping to contaminate public debate. To prevent this, it is crucial to develop deepfake social media messages detection systems. However, to the best of our knowledge no one has ever addressed the detection of machinegenerated texts on social networks like Twitter or Facebook. With the aim of helping the research in this detection field, we collected the first dataset of real deepfake tweets, Tweep- Fake. It is real in the sense that each deepfake tweet was actually posted on Twitter. We collected tweets from a total of 23 bots, imitating 17 human accounts. The bots are based on various generation techniques, i.e., Markov Chains, RNN, RNN+Markov, LSTM, GPT-2. We also randomly selected tweets from the humans imitated by the bots to have an overall balanced dataset of 25,572 tweets (half human and half bots generated). The dataset is publicly available on Kaggle. Lastly, we evaluated 13 deepfake text detection methods (based on various state-of-the-art approaches) to both demonstrate the challenges that Tweepfake poses and create a solid baseline of detection techniques. We hope that Tweep- Fake can offer the opportunity to tackle the deepfake detection on social media messages as well
Enriching Digital Libraries with Crowdsensed Data: Twitter Monitor and the SoBigData Ecosystem
No full textSoBigData is a Research Infrastructure (RI) aiming to provide an integrated ecosystem for ethic-sensitive scientific discoveries and advanced applications of social data mining. A key milestone of the project focuses on data, methods and results sharing, in order to ensure the reproducibility, review and re-use of scientific works. For this reason, the Digital Library paradigm is implemented within the RI, providing users with virtual environments where datasets, methods and results can be collected, maintained, managed and preserved, granting full documentation, access and the possibility to re-use. In this paper, we describe the results of our effort for integrating the Twitter Monitor, a tool for gathering messages from the Twitter Online Social Network, into the SoBigData RI. The Twitter Monitor provides a simple user interface, enabling researchers and stakeholders, without programming skills, to seamlessly (i) select relevant messages out of the huge Twitter stream by means of language, keyword, user tracking and geographical filters, (ii) store data on user personal Workspace, (iii) and publish them in the SoBigData Resource Catalogue, which implements all the aforementioned Digital Library features. Thanks to the seamless integration in the SoBigData RI, the Twitter Monitor allows researchers and stakeholders, belonging to different areas and having different backgrounds, to exploit the crowdsensing paradigm for enriching the SoBigData Digital Library. In this way, crowdsensing acquires the key features of openness, accessibility, interoperability and interdisciplinarity that characterize the Digital Libraries framework
Integration of External Data in Document Workflows via Web Services
No full textIn recent years the adoption of systems to automatically manage information assets in networked, collaborative ways has become vital for enterprises. In this work we focus on collaborative document management systems, where actual document development relies on accessing structured data in remote repositories. Forms within documents must be filled in with proper data items: An automated support to gather the required remote data may substantially improve the overall document management process. The Web Service technology is of fundamental help in this setting, because it lets remote data be exposed to potential clients according to standard interfaces. On the other hand, document management systems must be modified to provide this new kind of data access. This paper introduces a novel approach to support remote data integration in documents handled by Content Management Systems. According to the reported experience, such a data integration can be regarded as an enabling feature for a better exploitation of document workflow systems
Coordinated inauthentic behavior and information spreading on Twitter
No full textWe explore the effects of coordinated users (i.e., users characterized by an unexpected, suspicious, or exceptional similarity) in information spreading on Twitter by quantifying the efficacy of their tactics in deceiving feed algorithms to maximize information outreach. In particular, we investigate the behavior of coordinated accounts within a large set of retweet-based information cascades identifying key differences between coordinated and non-coordinated accounts in terms of position within the cascade, action delay and outreach. On average, coordinated accounts occupy higher positions of the information cascade (i.e., closer to the root), spread messages faster and involve a slightly higher number of users. When considering cascade metrics such as size, number of edges and height, we observe clear differences among information cascades that are associated to a systematically larger proportion of coordinated accounts, as confirmed by comparisons with statistical null models. To further characterize the activity of coordinated accounts we introduce two new measures capturing their infectivity within the information cascade (i.e., their ability to involve other users) and their interaction with non-coordinated accounts. Finally, we find that the interaction pattern between the two classes of users follows a saturation-like process. A larger-scale targeting of non-coordinated users does not require a larger amount of coordinated accounts after a threshold value similar to 50%, after which involving more coordinated accounts within a cascade yields a null marginal effect. Our results contribute to shed light on the role of coordinated accounts and their effect on information diffusion
- …
