Online Social Media (OSM) have been substantially transforming the process of spreading news, improving its speed, and reducing barriers toward reaching out to a broad audience. However, OSM are very limited in providing mechanisms to check the credibility of news propagated through their structure. The majority of studies on automatic fake news detection are restricted to English documents, with few works evaluating other languages, and none comparing language-independent characteristics. Moreover, the spreading of deceptive news tends to be a worldwide problem; therefore, this work evaluates textual features that are not tied to a specific language when describing textual data for detecting news. Corpora of news written in American English, Brazilian Portuguese, and Spanish were explored to study complexity, stylometric, and psychological text features. The extracted features support the detection of fake, legitimate, and satirical news. We compared four machine learning algorithms (k-Nearest Neighbors (k-NN), Support Vector Machine (SVM), Random Forest (RF), and Extreme Gradient Boosting (XGB)) to induce the detection model. Results show our proposed language-independent features are successful in describing fake, satirical, and legitimate news across three different languages, with an average detection accuracy of 85.3% with RF.
Online process mining refers to a class of techniques for analyzing in real-time event streams generated by the execution 5 of business processes. These techniques are crucial in the reactive monitoring of business processes, timely resource allocation and 6 detection/prevention of dysfunctional behavior. Many interesting advances have been made by the research community in recent 7 years, but there is no consensus on the exact set of properties these techniques have to achieve. This article fills the gap by identifying 8 a set of evaluation goals for online process mining and examining their fulfillment in the state of the art. We discuss parameters and 9 techniques regulating the balance between conflicting goals and outline research needed for their improvement. Concept drift detection 10 is crucial in this sense but, as demonstrated by our experiments, it is only partially supported by current solutions. Q1 11 Index Terms-Online process mining, event stream, requirements and goals, concept drift Ç 12 1 INTRODUCTION 13 P ROCESS Mining (PM) is a set of data science techniques 14 focused on the analysis of event logs [1]. Events are 15 recorded when executing a Business Process and collected 16 into cases, i.e., end to end sequences of events relevant to the 17 same process instance. Traditional PM algorithms were 18 designed to work offline, analyzing historical batches of logs 19 gathering the complete course of cases, if necessary with 20 multiple passes of analysis. This is, however, insufficient, 21 from a business standpoint, when the real-time assessment 22 of processes is crucial to timely manage resources and 23 quickly react to dysfunctional behaviors [2]. Today's fast-24 changing market requires systematic adjustments of pro-25 cesses in response to changes in the organization's operat-26 ing system or to trends emerging from the environment [3]. 27 Recently, the notion of online PM has emerged in reference 28 to analytics capable of handling real-time event streams [4], 29 [5]. An event stream differs from an event log because it is an 30 unbounded sequence of events ingested one-by-one and 31 allowing for limited actions in terms of iteration and memory 32 or time consumption [6].
Encoding methods affect the performance of process mining tasks but little work in the literature focused on quantifying their impact. In this paper, we compare 10 different encoding methods from three different families (trace replay and alignment, graph embeddings, and word embeddings) using measures to evaluate the overlaps in the feature space, the accuracy obtained, and the computational resources (time) consumed with a classification task. Across hundreds of event logs representing four variations of five scenarios and five anomalies, it was possible to identify the edge2vec method as the most accurate and effective in reducing class overlapping in the feature space.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.