Web objectionable text content detection using topic modeling technique

2013 ◽  
Vol 40 (15) ◽  
pp. 6094-6104 ◽  
Author(s):  
Jiangjiao Duan ◽  
Jianping Zeng
Keyword(s):  
Topic Modeling ◽  
Modeling Technique ◽  
Text Content

scholarly journals The recipes of Philosophy of Science: Characterizing the semantic structure of corpora by means of topic associative rules

PLoS ONE ◽  
2020 ◽  
Vol 15 (11) ◽  
pp. e0242353
Author(s):  
Christophe Malaterre ◽  
Jean-François Chartier ◽  
Francis Lareau
Keyword(s):  
Philosophy Of Science ◽  
Full Text ◽  
Topic Modeling ◽  
Semantic Content ◽  
Semantic Structure ◽  
Language Generation ◽  
Topological Features ◽  
Text Content ◽  
F Measure

Scientific articles have semantic contents that are usually quite specific to their disciplinary origins. To characterize such semantic contents, topic-modeling algorithms make it possible to identify topics that run throughout corpora. However, they remain limited when it comes to investigating the extent to which topics are jointly used together in specific documents and form particular associative patterns. Here, we propose to characterize such patterns through the identification of “topic associative rules” that describe how topics are associated within given sets of documents. As a case study, we use a corpus from a subfield of the humanities—the philosophy of science—consisting of the complete full-text content of one of its main journals: Philosophy of Science. On the basis of a pre-existing topic modeling, we develop a methodology with which we infer a set of 96 topic associative rules that characterize specific types of articles depending on how these articles combine topics in peculiar patterns. Such rules offer a finer-grained window onto the semantic content of the corpus and can be interpreted as “topical recipes” for distinct types of philosophy of science articles. Examining rule networks and rule predictive success for different article types, we find a positive correlation between topological features of rule networks (connectivity) and the reliability of rule predictions (as summarized by the F-measure). Topic associative rules thereby not only contribute to characterizing the semantic contents of corpora at a finer granularity than topic modeling, but may also help to classify documents or identify document types, for instance to improve natural language generation processes.

scholarly journals Topic Modeling Technique for Text Mining Over Biomedical Text Corpora Through Hybrid Inverse Documents Frequency and Fuzzy K-Means Clustering

IEEE Access ◽  
2019 ◽  
Vol 7 ◽  
pp. 146070-146080 ◽  
Author(s):  
Junaid Rashid ◽  
Syed Muhammad Adnan Shah ◽  
Aun Irtaza ◽  
Toqeer Mahmood ◽  
Muhammad Wasif Nisar ◽  
...  
Keyword(s):  
Text Mining ◽  
Topic Modeling ◽  
Modeling Technique ◽  
Biomedical Text ◽  
Text Corpora

Accounting for Accounting History: A topic modeling approach (1996–2015)

2017 ◽  
Vol 23 (1-2) ◽  
pp. 173-205 ◽  
Author(s):  
Paolo Ferri ◽  
Maria Lusiani ◽  
Luca Pareschi
Keyword(s):  
Topic Modeling ◽  
Modeling Technique ◽  
Accounting History ◽  
Modeling Approach ◽  
Research Areas ◽  
Theoretical Approaches ◽  
Relative Prevalence ◽  
Over Time

This article analyses all articles published in Accounting History using a topic modeling technique. Previous studies focus on the content of accounting history, but not how the field has evolved. The article complements prior assessments of the research published in Accounting History by providing measures of the relative prevalence of research areas and their evolution over time. The analysis offers insights into accounting history by refining previous categorisations, uncovering overlooked topic areas and substantiating trends, such as the demise of interest in the technical core of accounting in favour of more variegated and fragmented approaches. The findings are discussed in light of the claimed pluralisation of methodological and theoretical approaches in this field.

scholarly journals Pro/Anti-vaxxers in Brazil: a temporal analysis of COVID vaccination stance in Twitter

2021 ◽  
Author(s):  
Andre Mediote de Sousa ◽  
Karin Becker
Keyword(s):  
Federal Government ◽  
Topic Modeling ◽  
Temporal Analysis ◽  
Modeling Technique ◽  
Political Bias ◽  
Mandatory Vaccination

Collective imunization is critical to combat COVID, but a large portion of the population in many countries refuses to be vaccinated despite the availability of vaccines. We developed a temporal analysis of pro/against stances towards COVID vaccination in Brazil using Twitter. We summarized the main topics expressed by pro/anti-vaxxers using BERTopic, a dynamic topic modeling technique, and related them to events in the national scenario. The anti-vaxxers were prevalent throughout 2020, expressing concerns about mandatory vaccination with a strong political bias. The pro-vaxxer movement significantly increased by late 2020 with the begging of immunization and became prevalent in 2021. This group expresses joy and anxiety to get vaccinated and criticisms towards the Federal Government.

Topic-based content and sentiment analysis of Ebola virus on Twitter and in the news

2016 ◽  
Vol 42 (6) ◽  
pp. 763-781 ◽  
Author(s):  
Erin Hea-Jin Kim ◽  
Yoo Kyung Jeong ◽  
Yuyoung Kim ◽  
Keun Young Kang ◽  
Min Song
Keyword(s):  
Life Span ◽  
Sentiment Analysis ◽  
News Media ◽  
Topic Modeling ◽  
Ebola Virus ◽  
Modeling Technique ◽  
Entity Extraction ◽  
Query Term ◽  
Content Coverage ◽  
N Gram

The present study investigates topic coverage and sentiment dynamics of two different media sources, Twitter and news publications, on the hot health issue of Ebola. We conduct content and sentiment analysis by: (1) applying vocabulary control to collected datasets; (2) employing the n-gram LDA topic modeling technique; (3) adopting entity extraction and entity network; and (4) introducing the concept of topic-based sentiment scores. With the query term ‘Ebola’ or ‘Ebola virus’, we collected 16,189 news articles from 1006 different publications and 7,106,297 tweets with the Twitter stream API. The experiments indicate that topic coverage of Twitter is narrower and more blurry than that of the news media. In terms of sentiment dynamics, the life span and variance of sentiment on Twitter is shorter and smaller than in the news. In addition, we observe that news articles focus more on event-related entities such as person, organization and location, whereas Twitter covers more time-oriented entities. Based on the results, we report on the characteristics of Twitter and news media as two distinct news outlets in terms of content coverage and sentiment dynamics.

scholarly journals AN EFFICIENT NOVEL APPROACH FOR SOCIETAL COMMUNICATION OF Q&A COMMUNITY SYSTEM USING TOPIC MODELING TECHNIQUES

2021 ◽  
Vol 9 (1) ◽  
pp. 1270-1282
Author(s):  
Venkateswara Rao P, A.P Siva kumar
Keyword(s):  
Topic Modeling ◽  
Web Search ◽  
Key Factors ◽  
Public Preferences ◽  
Community System ◽  
Novel Approach ◽  
Word Context ◽  
The World ◽  
Questions And Answers ◽  
Text Content

The emerging trend in technical research is to use customer-generated data collected by community media to probe community opinion and scientific communication on employment and care issues. This review of the collected data, the launch of a question-and-answer social website, is a separate stack for exploring the key factors that influence public preferences for technical knowledge and opinions. by means of a web search engine, topic modeling, and regression data modeling, this study quantified the effect of the response textual and auxiliary functions on the number of votes received with the response. Compared to previous studies based on open estimates, the model results show that Quora users are more likely to only talk about technology. It can fail when the keywords in the query do not match the text content of large documents that contain relevant questions of existing methods, ie. CNNMF and NMF, as well as some restrictions are not enough. Also, users are often not experts and provide ambiguous queries leading to mixed results and encountering problems with existing methods. To address this problem, in this article we propose a Hadoop model, distributed using semantics, non-negative matrix factorization (HDiSANNMF), to find topics for short texts. It effectively incorporates the semantic correlations of the word context into the model, where the semantic connections between words and their context are learned by omitting the grammatical view of the corpus. The researchers are trying to reorganize the main results and present modern techniques for modeling distributed themes to address technologies and platforms with increasing attributes, as well as how much time and space it takes to generate the model. This document briefly describes the structure of public questions and answers around the world and tracks the development of the main topics Housing and employment opportunities for next generation technologies in the world in real time.

Sign in / Sign up

Export Citation Format

Share Document