scholarly journals Can Fake News Detection Models Maintain the Performance through Time? A Longitudinal Evaluation of Twitter Publications

Mathematics ◽  
2021 ◽  
Vol 9 (22) ◽  
pp. 2988
Author(s):  
Nuno Guimarães ◽  
Álvaro Figueira ◽  
Luís Torgo
Keyword(s):  
Social Networks ◽  
Negative Impact ◽  
Feature Model ◽  
Training Data ◽  
Fake News ◽  
Longitudinal Evaluation ◽  
Short Period ◽  
Testing Data ◽  
Using Data

The negative impact of false information on social networks is rapidly growing. Current research on the topic focused on the detection of fake news in a particular context or event (such as elections) or using data from a short period of time. Therefore, an evaluation of the current proposals in a long-term scenario where the topics discussed may change is lacking. In this work, we deviate from current approaches to the problem and instead focus on a longitudinal evaluation using social network publications spanning an 18-month period. We evaluate different combinations of features and supervised models in a long-term scenario where the training and testing data are ordered chronologically, and thus the robustness and stability of the models can be evaluated through time. We experimented with 3 different scenarios where the models are trained with 15-, 30-, and 60-day data periods. The results show that detection models trained with word-embedding features are the ones that perform better and are less likely to be affected by the change of topics (for example, the rise of COVID-19 conspiracy theories). Furthermore, the additional days of training data also increase the performance of the best feature/model combinations, although not very significantly (around 2%). The results presented in this paper build the foundations towards a more pragmatic approach to the evaluation of fake news detection models in social networks.

scholarly journals Perbandingan Akurasi, Recall, dan Presisi Klasifikasi pada Algoritma C4.5, Random Forest, SVM dan Naive Bayes

2021 ◽  
Vol 5 (2) ◽  
pp. 640
Author(s):  
Mulkan Azhari ◽  
Zakaria Situmorang ◽  
Rika Rosnelly
Keyword(s):  
Random Forest ◽  
Naive Bayes ◽  
Naïve Bayes ◽  
Training Data ◽  
Random Forest Algorithm ◽  
Svm Algorithm ◽  
Testing Data ◽  
C4.5 Algorithm ◽  
Bayes Algorithm ◽  
Using Data

In this study aims to compare the performance of several classification algorithms namely C4.5, Random Forest, SVM, and naive bayes. Research data in the form of JISC participant data amounting to 200 data. Training data amounted to 140 (70%) and testing data amounted to 60 (30%). Classification simulation using data mining tools in the form of rapidminer. The results showed that . In the C4.5 algorithm obtained accuracy of 86.67%. Random Forest algorithm obtained accuracy of 83.33%. In SVM algorithm obtained accuracy of 95%. Naive Bayes' algorithm obtained an accuracy of 86.67%. The highest algorithm accuracy is in SVM algorithm and the smallest is in random forest algorithm

Nexus Between Debt Financing and Market Value Moderated by Firm Size: Panel Data Evidence from India

2021 ◽  
pp. 097226292110526
Author(s):  
Rajesh Desai
Keyword(s):  
Firm Size ◽  
Firm Value ◽  
Small Firms ◽  
Negative Impact ◽  
Market Value ◽  
Debt Financing ◽  
Least Square ◽  
Short Term ◽  
Using Data

This study examines the effect of debt financing on market value of firm and evaluates the moderating effect of firm size on this relationship. Tobin’s Q and market-to-book value ratio are used as proxy for market value whereas long-term as well as short-term debt ratios are considered to indicate debt financing. Using data of 164 capital goods sector companies for 10 years (from 2010 to 2019), panel least square (PLS) regression with fixed and random effects (RE) model has been applied for data analysis. Based on findings, the study reports significant negative impact of borrowings (both long-term and short-term) on market value of selected companies. Further, the outcome of study confirms that firm size moderates the relationship between debt financing and firm value. The magnitude and significance of the effect of debt are stronger for small firms as compared to medium and large firms. Present verdicts will assist managers in designing capital structure policies by considering its impact on market value according to firm-size.

The value of social capital in farmland leasing relationships

2018 ◽  
Vol 78 (4) ◽  
pp. 489-496 ◽  
Author(s):  
Mykel R. Taylor ◽  
Allen M. Featherstone
Keyword(s):  
Social Capital ◽  
Design Methodology ◽  
Agricultural Land ◽  
Negative Impact ◽  
Content Type ◽  
Sample Average ◽  
Percent Increase ◽  
Rental Rate ◽  
Using Data

Purpose The purpose of this paper is to investigate the impacts of social capital on the rate at which agricultural land is rented between landowners and tenants using data from the state of Kansas. Design/methodology/approach A survey of tenants provides data on the rental rate of farmland as well as characteristics of the lease, the land, and the landowner. Findings Results support the hypothesis of a negative impact on rental rates from longer-term leasing relationships. The model estimates a 10.0 percent discount relative to market rates when the leasing relationship increases from 11 to 22 years. At the sample average of $64 per acre, this is a $10 per acre discount. Research limitations/implications Increased levels of social capital, as measured by the length of the leasing relationship between landowner and tenant, reduce the rental rate. A 10 percent increase in the number of years a parcel of land is leased to the same tenant will decrease the annual rental rate by 1 percent. Originality/value Research adds to the understanding of informal relationships underlying farmland leases. A large number of farmland tracts may turnover in the coming years. This turnover may affect the rental rates for tenants who have had long-term leasing relationships over time.

scholarly journals BOTS AMONG US: PREVALENCE, INFLUENCE, AND ROLES OF AUTOMATED ACCOUNTS IN THE GERMAN TWITTER FOLLOW NETWORK

2019 ◽  
Vol 2019 ◽  
Author(s):  
Felix Victor Münch ◽  
Cornelius Puschmann ◽  
Ben Thies ◽  
Axel Bruns
Keyword(s):  
Negative Impact ◽  
Fake News ◽  
National Scale ◽  
False Negatives ◽  
Graph Exploration ◽  
German Speaking ◽  
Potential Impact ◽  
The Cost ◽  
Term Monitoring

Social bots are undermining trust in social media. They spread low-credibility content, fake news, and spam. However, most research is based on bots that actively share links or keywords, rather than assessing the longer-term presence of bots as an integral part of platforms. To address this gap, we present what to our knowledge is the first study that assesses the prevalence, influence, and roles of automated accounts in a Twitter follow network on a national scale. This allows us to analyse the potential impact of bots beyond the context of single events and topics. To collect a follow network of the most central accounts in the German-speaking Twittersphere, we have adapted the rank-degree method, a graph exploration method that is able to identify the most influential spreaders within complex networks, as a data mining method using the cost-free standard Twitter API. To identify bots, we employ the Botometer API. Both methods combined allow us to localise bots within topical clusters, to estimate their potential influence, and to assess the roles of the most central bots. Our findings indicate that bots have only a low negative impact on the German-speaking Twittersphere. However, the most sophisticated bots will likely remain absent from our study, as false negatives. Similarly, trolls and semi-automated accounts necessitate further research. Our new sampling approach combined with Botometer is promising, for example, for Twitterspheres based on other languages. The study itself opens further avenues of enquiry, such as a long-term monitoring of automated accounts in the German Twittersphere.

scholarly journals Ranking Influential Nodes of Fake News Spreading on Mobile Social Networks

2021 ◽  
Vol 29 (4) ◽  
pp. 93-130
Author(s):  
Yunfei Xing ◽  
Xiwei Wang ◽  
Feng-Kwei Wang ◽  
Yang Shi ◽  
Wu He ◽  
...  
Keyword(s):  
Social Networks ◽  
Information Entropy ◽  
Dynamic Process ◽  
Negative Impact ◽  
Entropy Theory ◽  
Direct Influence ◽  
Mobile Social Networks ◽  
Fake News ◽  
Two Factors ◽  
Influential Nodes

Online fake news can generate a negative impact on both users and society. Due to the concerns with spread of fake news and misinformation, assessing the network influence of online users has become an important issue. This study quantifies the influence of nodes by proposing an algorithm based on information entropy theory. Dynamic process of influence of nodes is characterized on mobile social networks (MSNs). Weibo (i.e., the Chinese version of microblogging) users are chosen to build the real network and quantified influence of them is analyzed according to the model proposed in this paper. MATLAB is employed to simulate and validate the model. Results show the comprehensive influence of nodes increases with the rise of two factors: the number of nodes connected to them and the frequency of their interaction. Indirect influence of nodes becomes stronger than direct influence when the network scope rises. This study can help relevant organizations effectively oversee the spread of online fake news on MSNs.

scholarly journals MODELING OF SOLAR RADIATION WITH A NEURAL NETWORK

2018 ◽  
Vol 24 (3) ◽  
Author(s):  
VALENTIN STOYANOV ◽  
IVAYLO STOYANOV ◽  
TEODOR ILIEV
Keyword(s):  
Neural Network ◽  
Solar Radiation ◽  
Training Data ◽  
Training Algorithms ◽  
Data Set ◽  
Tilted Surface ◽  
Using Data ◽  
Artificial Neural Network Ann ◽  
Hidden Neurons

<p>Modeling of solar radiation with neural network could be used for real-time calculations of the radiation on tilted surfaces with different orientations. In the artificial neural network (ANN), latitude, day of the year, slope, surface azimuth and average daily radiation on horizontal surface are inputs, and average daily radiation on tilted surface of definite orientation is output. The possible ANN structure, the size of training data set, the number of hidden neurons, and the type of training algorithms were analyzed in order to identify the most appropriate model. The same ANN structure was trained and tested using data generated from the Klein and Theilacker model and long-term measurements. Reasonable accuracy was obtained for all predictions for practical need.</p>

scholarly journals MODELING OF SOLAR RADIATION WITH A NEURAL NETWORK

2018 ◽  
Vol 24 (3) ◽  
pp. 45-50
Author(s):  
VALENTIN STOYANOV ◽  
IVAYLO STOYANOV ◽  
TEODOR ILIEV
Keyword(s):  
Neural Network ◽  
Solar Radiation ◽  
Training Data ◽  
Training Algorithms ◽  
Data Set ◽  
Tilted Surface ◽  
Using Data ◽  
Artificial Neural Network Ann ◽  
Hidden Neurons

Modeling of solar radiation with neural network could be used for real-time calculations of the radiation on tilted surfaces with different orientations. In the artificial neural network (ANN), latitude, day of the year, slope, surface azimuth and average daily radiation on horizontal surface are inputs, and average daily radiation on tilted surface of definite orientation is output. The possible ANN structure, the size of training data set, the number of hidden neurons, and the type of training algorithms were analyzed in order to identify the most appropriate model. The same ANN structure was trained and tested using data generated from the Klein and Theilacker model and long-term measurements. Reasonable accuracy was obtained for all predictions for practical need.

Coming Together Through Falling Apart

2019 ◽  
Vol 40 (2) ◽  
pp. 151-168
Author(s):  
Heather Churchill ◽  
Jeremy M. Ridenour
Keyword(s):  
Data Analysis ◽  
Qualitative Data ◽  
Structural Changes ◽  
Systematic Approach ◽  
Qualitative Approach ◽  
Psychological Assessments ◽  
Transference And Countertransference ◽  
Testing Data ◽  
Psychoanalytically Oriented Psychotherapy

Abstract. Assessing change during long-term psychotherapy can be a challenging and uncertain task. Psychological assessments can be a valuable tool and can offer a perspective from outside the therapy dyad, independent of the powerful and distorting influences of transference and countertransference. Subtle structural changes that may not yet have manifested behaviorally can also be assessed. However, it can be difficult to find a balance between a rigorous, systematic approach to data, while also allowing for the richness of the patient’s internal world to emerge. In this article, the authors discuss a primarily qualitative approach to the data and demonstrate the ways in which this kind of approach can deepen the understanding of the more subtle or complex changes a particular patient is undergoing while in treatment, as well as provide more detail about the nature of an individual’s internal world. The authors also outline several developmental frameworks that focus on the ways a patient constructs their reality and can guide the interpretation of qualitative data. The authors then analyze testing data from a patient in long-term psychoanalytically oriented psychotherapy in order to demonstrate an approach to data analysis and to show an example of how change can unfold over long-term treatments.

Analysis of Impacts Malaysia's Territorial Limitation (Lockdown) Policy on Indonesia-Malaysia Border Area, Sanggau District

2020 ◽  
Vol 11 (1) ◽  
pp. 156-156
Author(s):  
Ira Patriani
Keyword(s):  
Social Issues ◽  
Negative Impact ◽  
Research Approach ◽  
Border Area ◽  
Border Areas ◽  
The Government ◽  
Using Data ◽  
Virus Spreading ◽  
Kinship Ties ◽  
Collection Methods

Border areas, is one of affected area on COVID_19 this present. Many of people cn not go out as usually, adding almost each country has to implement their territorial limitation (lockdown policy) to minimalize this virus spreading. One of Malaysia State, where very close and get direct border with Indonesia. This research took place at Sanggau District, Entikong, Gun Tembawang Village.The research approach used is qualitative, using data collection methods in the form of interviews, observations, and documentation supported by interviews with the theoretical approach to the negative and positive aspects on policy implementation. Research results, The results stated that the lockdown activities of Malaysia which were affected by the corona virus outbreak needed to be carried out in an effort to minimize the spread of the virus outbreak. Although of course it has a negative impact on the country's economic structure, social issues and other sector. In implementing this lockdown, there is a need for cooperation between the government and the community as well as an agreement with neighboring countries in terms of the mobility of residents closest to each other's territory on exemptions in order to realize social welfare and public health without limiting the origin of the state, religion, community and profession. Especially in border areas where mobility and kinship ties have always been closer than in other regions. Keywords: Border area, lockdown policy, covid_19

An Analog Circuit Fault Diagnosis Approach Based on Wavelet-based fractal analysis and Multiple Kernel SVM

2020 ◽  
Vol 13 ◽  
Author(s):  
Jianfeng Jiang
Keyword(s):  
Fault Diagnosis ◽  
Fractal Analysis ◽  
Analog Circuit ◽  
Training Data ◽  
Support Vector ◽  
Pass Filter ◽  
Multiple Kernel ◽  
Testing Data ◽  
Circuit Fault Diagnosis ◽  
Diagnosis Approach

Objective: In order to diagnose the analog circuit fault correctly, an analog circuit fault diagnosis approach on basis of wavelet-based fractal analysis and multiple kernel support vector machine (MKSVM) is presented in the paper. Methods: Time responses of the circuit under different faults are measured, and then wavelet-based fractal analysis is used to process the collected time responses for the purpose of generating features for the signals. Kernel principal component analysis (KPCA) is applied to reduce the features’ dimensionality. Afterwards, features are divided into training data and testing data. MKSVM with its multiple parameters optimized by chaos particle swarm optimization (CPSO) algorithm is utilized to construct an analog circuit fault diagnosis model based on the testing data. Results: The proposed analog diagnosis approach is revealed by a four opamp biquad high-pass filter fault diagnosis simulation. Conclusion: The approach outperforms other commonly used methods in the comparisons.

Sign in / Sign up

Export Citation Format

Share Document