scholarly journals A Herd Effect Detection Method Based on Text Features

2021 ◽  
Author(s):  
Tingzhen Liu ◽  
Tong Zhou ◽  
Yuxin Shi ◽  
Siyuan Liu ◽  
Jin Gao
Keyword(s):  
Social Network ◽  
Social Network Analysis ◽  
Network Analysis ◽  
Language Processing ◽  
Detection Method ◽  
Graph Structure ◽  
Mining Method ◽  
Herd Effect ◽  
Text Features ◽  
Unsupervised Data Mining

The herd effect is a common phenomenon in social society. The detection of this phenomenon is of great significance in many tasks based on social network analysis such as recommendation. However, the research on social network and natural language processing seldom focuses on this issue. In this paper, we propose an unsupervised data mining method to detect herding in social networks. Taking shopping review as an example, our algorithm can identify other reviews which are affected by some previous reviews and detect a herd effect chain. From the overall perspective, the cross effects of all views form the herd effect graph. This algorithm can be widely used in various social network analysis methods through graph structure, which provides new useful features for many tasks.

scholarly journals Τεχνικές αναλύσης κοινωνικών δικτύων, με έμφαση σε γράφους εμπιστοσύνης

2015 ◽  
Author(s):  
Αθανάσιος Παπαοικονόμου
Keyword(s):  
Deep Learning ◽  
Natural Language Processing ◽  
Social Network ◽  
Social Network Analysis ◽  
Network Analysis ◽  
Natural Language ◽  
Language Processing

Η παρούσα διατριβή προτείνει τεχνικές για την ανάλυση κοινωνικών δικτύων δίνοντας ιδιαίτερη έμφαση σε δίκτυα στα οποία οι χρήστες μπορούν να εκφράζουν εμπιστοσύνη ή δυσπιστία μεταξύ τους. Η ανάλυση τέτοιων γράφων εμπιστοσύνης είναι ένα ενδιαφέρον πρόβλημα με ευρύ φάσμα εφαρμογών όπως η ανάλυση γεωπολιτικών σχέσεων και η εύρεση κοινοτήτων χρηστών. Στα πρώτα τρία κεφάλαια εξετάζεται το πρόβλημα της πρόβλεψης της προδιάθεσης ενός χρήστη για έναν άλλο, αντλώντας τεχνικές από τρεις διαφορετικούς τομείς. Αρχικά, χρησιμοποιούνται κλασικές και διαδεδομένες τεχνικές από τον χώρο της Ανάλυσης Κοινωνικών Δικτύων (Social Network Analysis) με σκοπό να ερευνηθούν οι μηχανισμοί διάδοσης θετικών και αρνητικών απόψεων στο δίκτυο. Έπειτα, ενσωματώνουμε τεχνικές από τον τομέα της Βιοστατιστικής, ώστε να αναλύσουμε μεγάλα κοινωνικά δίκτυα από μικροσκοπική σκοπιά. Στη συνέχεια, με χρήση τεχνικών deep learning δείχνουμε πως είναι δυνατόν να "κατασκευαστεί" ένας γράφος εμπιστοσύνης αξιοποιώντας δεδομένα φαινομενικά άσχετα με αυτόν τον σκοπό, όπως οι κριτικές των χρηστών για διάφορα προϊόντα. Στο τελευταίο κεφάλαιο, παρουσιάζουμε έναν αλγόριθμο εύρεσης κοινοτήτων σε κοινωνικά δίκτυα, βασιζόμενοι σε πρόσφατες προόδους στον τομέα της Ανάλυσης Φυσικής Γλώσσας (Natural Language Processing). Η σειρά των κεφαλαίων αποτυπώνει την χρονική σειρά των πειραμάτων που εφάρμοσα αλλά κάθε κεφάλαιο είναι γραμμένο ώστε να μην έχει σημαντικές συσχετίσεις με τα προηγούμενα και έτσι να μπορεί να διαβαστεί αυτόνομα

scholarly journals Student’s Performance Prediction Using Hybrid Optimization Algorithm-based Map Reduce Framework

2021 ◽  
Author(s):  
Dipak Bhosale ◽  
Mahendra Sawane
Keyword(s):  
Social Network ◽  
Social Network Analysis ◽  
Network Analysis ◽  
Student Performance ◽  
Language Processing ◽  
Learning Analytics ◽  
Concept Maps ◽  
Machine Learning Techniques ◽  
Planning And Scheduling ◽  
Student’S Performance

Learning analytics (LA) is a growing research area, which aims at selecting, analyzing and reporting student data (in their interaction with the online learning environment), finding patterns in student behaviour, displaying relevant information in suggestive formats; the end goal is the prediction of student performance, the optimization of the educational platform and the implementation of personalized interventions. According to the Society of Learning Analytics Research1, LA can be defined as "the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimizing learning and the environments in which it occurs". The topic is highly interdisciplinary, including machine learning techniques, educational data mining, statistical analysis, social network analysis, natural language processing, but also knowledge from learning sciences, pedagogy and sociology; up-to-date overviews of the area are provided in. Various educational tasks can be supported by learning analytics, as identified in analysis and visualization of data; providing feedback for supporting instructors; providing recommendations for students; predicting student's performance; student modelling; detecting undesirable student behaviours; grouping students; social network analysis; developing concept maps; constructing courseware; planning and scheduling. Similarly, seven main objectives of learning analytics are summarized in: monitoring and analysis; prediction and intervention; tutoring and mentoring; assessment and feedback; adaptation; personalization and recommendation; reflection.

scholarly journals Multimodal Social Network Dataset Based on Goldbach Conjecture Proved Event in Zhihu

2021 ◽  
Author(s):  
Tingzhen Liu ◽  
Shijie Geng ◽  
Zhiquan Huang ◽  
Senxin Wu ◽  
Zixi Wang
Keyword(s):  
Natural Language Processing ◽  
Social Network ◽  
Social Network Analysis ◽  
Network Analysis ◽  
Natural Language ◽  
Language Processing ◽  
Dynamic Features ◽  
Fine Grained ◽  
Before And After ◽  
The Impact

At the end of 2018, a high school student asked a question in Zhihu community, claiming that he had proved Goldbach's conjecture. The problem caused an explosive reaction and a large number of users participated in the discussion. And has caused the widespread influence. On January 1, 2019, the questioner issued his "proof". His proof was soon proved wrong. The heated discussion caused by the incident contains a lot of information of social science analysis value. Therefore, we follow up the event in the first time and build a time series dataset for the event. Taking the questioner's "proof" as the dividing line, all the answers, comments, sub comments and user information of writing these texts before and after two days were recorded. This series of temporal information can reflect the dynamic features of the interaction between user opinions, and the impact of exogenous shocks (proof release) on community opinions. The dataset can be used not only for the demonstration of various social network analysis algorithms, but also for a series of natural language processing tasks such as fine-grained sentiment analysis for long texts, as well as multimodal tasks combining natural language processing and social network analysis. This paper introduces the characteristics and structure of the dataset, shows the visualization effect of social network, and uses the dataset train the benchmark model of emotion analysis.

scholarly journals Keyword extraction for film reviews based on social network analysis and natural language technology

2020 ◽  
Vol 189 ◽  
pp. 03019
Author(s):  
Quan Yanan ◽  
Tan Fuqiang
Keyword(s):  
Social Network ◽  
Social Network Analysis ◽  
Network Analysis ◽  
Natural Language ◽  
Language Processing ◽  
Semantic Analysis ◽  
Semantic Network ◽  
Keyword Extraction ◽  
Main Stream ◽  
Language Technology

At present, there are many movie reviews appear on main stream websites, and these evaluations are quite different to the same movie. As a customer, how to choose your favorite movie and television program? To solve this problem, this study attempts to use the semantic analysis of word vectors (Word2vec) semantic analysis in machine learning as a research tool to mine a large number of movie reviews. The research shows that most movie reviews have a certain theme cohesion and their semantic network has quite connected. Through the use of social network analysis and the use of Word2vec word vector technology in natural language processing, it is possible to present a streamlined movie review based on movie review network semantics and keyword extraction, thus helping to select the favorite movie review.

scholarly journals Opinion-based Homogeneity on YouTube : Combining Sentiment and Social Network Analysis

2020 ◽  
Vol 2 (1) ◽  
pp. 81-108 ◽  
Author(s):  
Daniel Röchert ◽  
German Neubaum ◽  
Björn Ross ◽  
Florian Brachten ◽  
Stefan Stieglitz
Keyword(s):  
Social Network ◽  
Social Network Analysis ◽  
Network Analysis ◽  
Language Processing ◽  
Online Communication ◽  
Future Research ◽  
Political Issues ◽  
Online Networks ◽  
Same Sex Couples ◽  
Political Discourses

Abstract When addressing public concerns such as the existence of politically like-minded communication spaces in social media, analyses of complex political discourses are met with increasing methodological challenges to process communication data properly. To address the extent of political like-mindedness in online communication, we argue that it is necessary to focus not only on ideological homogeneity in online environments, but also on the extent to which specific political questions are discussed in a uniform manner. This study proposes an innovative combination of computational methods, including natural language processing and social network analysis, that serves as a model for future research examining the evolution of opinion climates in online networks. Data were gathered on YouTube, enabling the assessment of users’ expressed opinions on three political issues (i.e., adoption rights for same-sex couples, headscarf rights, and climate change). Challenging widely held assumptions on discursive homogeneity online, the results provide evidence for a moderate level of connections between dissimilar YouTube comments but few connections between agreeing comments. The findings are discussed in light of current computational communication research and the vigorous debate on the prevalence of like-mindedness in online networks.

scholarly journals Opinion-based Homogeneity on YouTube: Combining Sentiment and Social Network Analysis

2019 ◽  
Author(s):  
Daniel Röchert ◽  
German Neubaum ◽  
Björn Ross ◽  
Florian Brachten ◽  
Stefan Stieglitz
Keyword(s):  
Social Network ◽  
Social Network Analysis ◽  
Network Analysis ◽  
Language Processing ◽  
Online Communication ◽  
Future Research ◽  
Political Issues ◽  
Online Networks ◽  
Same Sex Couples ◽  
Political Discourses

When addressing public concerns such as the existence of politically like-minded communication spaces in social media, analyses of complex political discourses are met with increasing methodological challenges to process communication data properly. To address the extent of political like-mindedness in online communication, we argue that it is necessary to focus not only on ideological homogeneity in online environments, but also on the extent to which specific political questions are discussed in a uniform manner. This study proposes an innovative combination of computational methods, including natural language processing and social network analysis, that serves as a model for future research examining the evolution of opinion climates in online networks. Data were gathered on YouTube, enabling the assessment of users’ expressed opinions on three political issues (i.e., adoption rights for same-sex couples, headscarf rights, and climate change). Challenging widely held assumptions on discursive homogeneity online, the results provide evidence for a moderate level of connections between dissimilar YouTube comments but few connections between agreeing comments. The findings are discussed in light of current computational communication research and the vigorous debate on the prevalence of like-mindedness in online networks.

scholarly journals Citizen participation in Twitter: Anti-vaccine controversies in times of COVID-19

Comunicar ◽  
2021 ◽  
Vol 29 (69) ◽  
Author(s):  
Rafael Carrasco-Polaino ◽  
Miguel-Ángel Martín-Cárdaba ◽  
Ernesto Villar-Cirujano
Keyword(s):  
Social Network ◽  
Social Network Analysis ◽  
Network Analysis ◽  
Citizen Participation ◽  
Language Processing ◽  
The Other ◽  
Citizen Engagement ◽  
Redes Sociales ◽  
Other Hand ◽  
First Time

Twitter has transformed into one of the main platforms for citizen engagement today. However, even though previous studies have focused on opinions about vaccines in general or about specific vaccines, opinions towards COVID-19 vaccines on Twitter have not been researched to date. The objective of this research is, by using social network analysis and language processing tools, to examine the degree to which the opinions and interactions present on Twitter are favorable or unfavorable towards the main COVID-19 vaccines. In addition, the relevance of each of the vaccines is studied, as well as their level of controversy. Likewise, the present study investigates, for the first time, the conversation from different perspectives including the content and also the participants, by analyzing in detail the verified accounts and using tools for the detection of bots. In global terms, the results from verified accounts show a moderate favorability towards the COVID-19 vaccines, the most accepted being those of Oxford-AstraZeneca, Pfizer, Moderna, and Sputnik V. On the other hand, the vaccine that attracts the most attention is the Russian Sputnik V, which is also the most controversial, behind those developed in China. Finally, verified users are shown to be relevant agents in the conversation due to their greater capacity for dissemination and reach, while the presence of bots is practically non-existent. Twitter se ha transformado en una de las principales plataformas de participación ciudadana hoy en día. Sin embargo, aun cuando estudios similares previos se han centrado en la opinión sobre las vacunas en general o sobre otras vacunas específicas, hasta la fecha no se han investigado las opiniones hacia las vacunas contra la COVID-19 en Twitter. El objetivo de esta investigación es, mediante el uso de herramientas de análisis de redes sociales y de herramientas de procesamiento del lenguaje, examinar el grado en el que las opiniones e interacciones presentes en Twitter son favorables o no hacia las principales vacunas de la COVID-19. Además, se estudia la relevancia de cada una de las principales vacunas, así como su nivel de controversia (polemicidad). Igualmente, el presente estudio investiga por primera vez la conversación no solo desde el punto de vista del contenido, sino también de los participantes que la integran, analizando en detalle las cuentas verificadas y empleando herramientas para la detección de bots. En términos globales, los resultados muestran una moderada favorabilidad hacia las vacunas de la COVID-19, siendo las más aceptadas las de Oxford-AstraZeneca, Pfizer y Moderna, y la de Sputnik V en el caso concreto de las cuentas verificadas. Por otro lado, la vacuna que más atención acapara es la rusa Sputnik V, que es además la más polémica por detrás de las de origen chino. Por último, los usuarios verificados se muestran como agentes relevantes de la conversación por su mayor capacidad de difusión y alcance, mientras que la presencia de bots es prácticamente inexistente.

Sign in / Sign up

Export Citation Format

Share Document