Social Bots Detection via Fusing BERT and Graph Convolutional Networks

The online social media ecosystem is becoming more and more confused because of more and more fake information and the social media of malicious users’ fake content; at the same time, unspeakable pain has been brought to mankind. Social robot detection uses supervised classification based on artificial feature extraction. However, user privacy is also involved in using these methods, and the hidden feature information is also ignored, such as semi-supervised algorithms with low utilization rates and graph features. In this work, we symmetrically combine BERT and GCN (Graph Convolutional Network, GCN) and propose a novel model that combines large scale pretraining and transductive learning for social robot detection, BGSRD. BGSRD constructs a heterogeneous graph over the dataset and represents Twitter as nodes using BERT representations. Corpus learning via text graph convolution network is a single text graph, which is mainly built for corpus-based on word co-occurrence and document word relationship. BERT and GCN modules can be jointly trained in BGSRD to achieve the best of merit, training data and unlabeled test data can spread label influence through graph convolution and can be carried out in the large-scale pre-training of massive raw data and the transduction learning of joint learning representation. The experiment shows that a better performance can also be achieved by BGSRD on a wide range of social robot detection datasets.

TV Show Dramaturgy as an Analogue of the Audiovisual Syntax

The dramaturgy of a television work as structure-forming for the format basis of a television periodical is examined. The purpose of the study is to explain the fundamental features of a TV program as an integral artistic phenomenon, embodied in many variations based on a single format model. The task is to clarify the problem of the effective functioning of syntactic rules in the absence of basic linguistic units, which is characteristic of the audiovisual language of screen shows, that has been actualized by domestic and Western media researchers. A general thesis is put forward about the polymodality of an audiovisual work as a single text, the unfolding of which occurs according to pre-established schemes that regulate the effect of individual elements on the viewer's perception. The dramatic approach is considered in the discourse of the global iconic turn from the verbal culture to the visual culture, there are multi-branch attempts to identify possible basic units of the new combinatorial sign system. Syntactic constructions, defined as a chain of episodes, show signs of fragmentation, mosaicism, intertextuality, and other criteria of postmodern aesthetics. The dualism of screen attractions as aggressively influencing instruments and verbal narrative, their mutual influence and dramatic significance are emphasized. The concepts of syntactic uniformity, the constancy of the syntactic structure chosen for a given format, without the possibility of breaking it in variations of editions are considered.

Chinese Gretchen in Russian Literature: on the Genesis and Attribution of M. Shkapskaya’s Poetry Book Tsa-Tsa-Tsa

This article examines in detail Maria Shkapskayas poetry book Tsa-Tsa- Tsa (1923) and its handwritten genesis. It explains the role and significance of ancient Chinese poetry for this literary piece of work. The problem is to attribute the texts that make up the book and find out their translated or stylized basis. The general thesis is that all the poetic texts of the book are translations: the names of Tao-Yuan-Ming, Du Fu, and Bo-Juyi indicated by Shkapskaya in the manuscripts are reported. One of the texts in the book is attributed as the Sixth Poem from the Shi ju gu shi ( Nineteen Ancient Poems ). The removal of the names of Chinese authors (not only in the book published in 1923 but also in the manuscript of 1921) and the alignment of the thematic word series silk, crane, thousand, spring that organize the book into a single text indicate a tendency to blur the border of the own-alien text (even though the book was treated by the author as translation from the Chinese, in autobiographies and correspondence). This trend leads to the appearance of a central artistic image of the book (it is a feature of M. Shkapskayas poetic books). It is the image of a lonely, longing woman. The mention of the spinning wheel connects this image with the popular (especially in Western European literature) image of Gretchen. This way the poetry book Tsa-Tsa-Tsa goes beyond the narrowly translated work and reveals some features of chronologically later literary trends (such as postmodernism and metapoesis).

Extraction of the Relations between Significant Pharmacological Entities in Russian-Language Internet Reviews on Medications

Nowadays, an analysis of virtual media to predict society’s reaction to any events or processes is a task of great relevance. Especially it concerns meaningful information on healthcare problems. Internet sources contain a large amount of pharmacologically meaningful information useful for pharmacovigilance purposes and repurposing drug use. An analysis of such a scale of information demands developing the methods that require the creation of a corpus with labeled relations among entities. Before, there have been no such Russian language datasets. This paper considers the first Russian language dataset where labeled entity pairs are divided into multiple contexts within a single text (by used drugs, by different users, by the cases of use, etc.), and a method based on the XLM-RoBERTa language model, previously trained on medical texts to evaluate the state-of-the-art accuracy for the task of indication of the four types of relationships among entities: ADR–Drugname, Drugname–Diseasename, Drugname–SourceInfoDrug, Diseasename–Indication. As shown based on the presented dataset from the Russian Drug Review Corpus, the developed method achieves the F1-score of 81.2% (obtained using cross-validation and averaged for the four types of relationships), which is 7.8% higher than the basic classifiers.

Bert-Enhanced Text Graph Neural Network for Classification

Text classification is a fundamental research direction, aims to assign tags to text units. Recently, graph neural networks (GNN) have exhibited some excellent properties in textual information processing. Furthermore, the pre-trained language model also realized promising effects in many tasks. However, many text processing methods cannot model a single text unit’s structure or ignore the semantic features. To solve these problems and comprehensively utilize the text’s structure information and semantic information, we propose a Bert-Enhanced text Graph Neural Network model (BEGNN). For each text, we construct a text graph separately according to the co-occurrence relationship of words and use GNN to extract text features. Moreover, we employ Bert to extract semantic features. The former part can take into account the structural information, and the latter can focus on modeling the semantic information. Finally, we interact and aggregate these two features of different granularity to get a more effective representation. Experiments on standard datasets demonstrate the effectiveness of BEGNN.

Engaging in dialogue during collaborative writing: The role of affective, cognitive, and social engagement

Collaborative writing (CW) involves two or more students writing a single text together. Previous studies mainly focused on students’ cognitive engagement in CW and investigated their attention to various language-related problems during task interaction. However, little CW research to date has considered that engagement in language-related discussions can manifest from three dimensions: cognitive, social, and affective. Focusing on the multidimensional characteristics of engagement, this study investigated how Russian learners’ social and affective reactions influence their focus on language use while they completed a CW task. Drawing on Svalberg’s framework of engagement with language to identify the three dimensions of engagement, I conducted a mixed-method approach towards analysing the audio-recorded collaborative dialogues by three student pairs ( n = 6), along with a qualitative analysis of their responses to a five-point Likert scale questionnaire. The analyses showed that when learners were interactive and viewed the activity as useful, they noticed many linguistic problems and elaborated on them. In contrast, when learners demonstrated social disengagement and perceived disadvantages from CW, they were likely to withdraw their attention from resolving the language issues they encountered. These findings indicate the complex and dynamic nature of task engagement. They can provide second language (L2) teachers with an in-depth understanding of how to fully engage students in instructional activities to better foster their L2 learning.

Scheduling and Control of Queueing Networks

Applications of queueing network models have multiplied in the last generation, including scheduling of large manufacturing systems, control of patient flow in health systems, load balancing in cloud computing, and matching in ride sharing. These problems are too large and complex for exact solution, but their scale allows approximation. This book is the first comprehensive treatment of fluid scaling, diffusion scaling, and many-server scaling in a single text presented at a level suitable for graduate students. Fluid scaling is used to verify stability, in particular treating max weight policies, and to study optimal control of transient queueing networks. Diffusion scaling is used to control systems in balanced heavy traffic, by solving for optimal scheduling, admission control, and routing in Brownian networks. Many-server scaling is studied in the quality and efficiency driven Halfin–Whitt regime and applied to load balancing in the supermarket model and to bipartite matching in ride-sharing applications.

E.K. Pekarsky’s dictionary of the yakut language as a kind of special supertext

This article studies E.K. Pekarsky’s Dictionary of the Yakut language within the scientific problem related to a supertext. Linguistic and literary works consider supertext as a multidimensional, cross-temporal, cross-personal, and polygenre phenomenon within the framework of the interpreted integrity of a number of independent texts; and integrity as a paradigmatic component is a constitutive part of the supertext. In supertexts, the textual component forms intertextual links and is revealed in the textual concept sphere, and simultaneously forms supertext links, creating a single text space.

Interactive Exploration of the Readability of Science Authors

Here we present a new text analysis tool that consists of a text analysis service and an author search service. These services were created by using or extending many existing Free and Open Source tools, including streamlit, requests, WordCloud, TextStat, and The Natural Language Tool Kit. The tool has the capability to retrieve journal hosting links and journal article content from APIs and journal hosting websites. Together, these services allow the user to review the complexity of a scientist’s published work relative to other online-based text repositories. Rather than providing feedback as to the complexity of a single text as previous tools have done, the tool presented here shows the relative complexity across many texts from the same author, while also comparing the readability of the author’s body of work to a variety of other scientific and lay text types. The goal of this work is to apply a more data-driven approach that provides established academic authors with statistical insights into their body of published peer reviewed work. By monitoring these readability metrics, scientists may be able to cater their writing to reach broader audiences, contributing to an improved global communication and understanding of complex topics.