Difference between revisions of "Supervised Machine Learning for Fake News Detection"
Adelo Vieira (talk | contribs) (→Gofaas Web App) |
Adelo Vieira (talk | contribs) (→Summary of Results) (Tag: Visual edit) |
||
Line 27: | Line 27: | ||
===Procedure=== | ===Procedure=== | ||
− | * The Dataset | + | *The Dataset |
− | * Splitting the data into Train and Test data | + | *Splitting the data into Train and Test data |
− | * Cleaning the data | + | *Cleaning the data |
− | * Building the Document-Term Matrix | + | *Building the Document-Term Matrix |
− | * Model Building | + | *Model Building |
− | * Cross validation | + | *Cross validation |
− | * Making predictions from the model created and displaying a Confusion matrix | + | *Making predictions from the model created and displaying a Confusion matrix |
<br /> | <br /> | ||
Line 50: | Line 50: | ||
{| class="wikitable" | {| class="wikitable" | ||
− | ! rowspan=" | + | ! rowspan="4" |[[Establishing an authenticity of sport news by Machine Learning Models#Algorithms|Algorithms]] |
− | ! rowspan=" | + | ! rowspan="4" |Author |
− | ! rowspan=" | + | ! rowspan="4" |Package |
− | ! rowspan=" | + | ! rowspan="4" |Keyword |
− | ! colspan=" | + | ! colspan="8" |Accuracy |
− | |||
|- | |- | ||
! colspan="2" |[[Establishing an authenticity of sport news by Machine Learning Models#Description of the Kaggle fake news dataset|Kaggle fake news dataset]] | ! colspan="2" |[[Establishing an authenticity of sport news by Machine Learning Models#Description of the Kaggle fake news dataset|Kaggle fake news dataset]] | ||
− | + | 20,800 rows | |
! colspan="2" |[[Establishing an authenticity of sport news by Machine Learning Models#Description of the Fake news Detector dataset|Fake news Detector dataset]] | ! colspan="2" |[[Establishing an authenticity of sport news by Machine Learning Models#Description of the Fake news Detector dataset|Fake news Detector dataset]] | ||
− | ! colspan=" | + | 10,000 rows |
+ | ! colspan="4" |Gofaaas Fake News Dataset | ||
+ | 500 rows | ||
+ | |- | ||
+ | ! rowspan="2" |Test data (70% of the dataset) | ||
+ | ! rowspan="2" |Cross validation | ||
+ | ! rowspan="2" |Test data (70% of the dataset) | ||
+ | ! rowspan="2" |Cross validation | ||
+ | ! rowspan="2" |Test data (70% of the dataset) | ||
+ | ! rowspan="2" |Cross validation | ||
+ | ! rowspan="2" |Using the Kaggle Model^ | ||
+ | ! rowspan="2" |Using the Detector Model^^ | ||
|- | |- | ||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
|- | |- | ||
|[[Establishing an authenticity of sport news by Machine Learning Models#Extreme Gradient Boosting|Naive Bayes]] | |[[Establishing an authenticity of sport news by Machine Learning Models#Extreme Gradient Boosting|Naive Bayes]] | ||
|[[wikipedia:Thomas_Bayes|Bayes, Thomas]] | |[[wikipedia:Thomas_Bayes|Bayes, Thomas]] | ||
− | | | + | |[https://cran.r-project.org/web/packages/e1071/index.html e1071] |
|NB* | |NB* | ||
− | |||
|[[Establishing an authenticity of sport news by Machine Learning Models#fnd_naivebayes|<math>{\color{blue}76.41%}</math>]] | |[[Establishing an authenticity of sport news by Machine Learning Models#fnd_naivebayes|<math>{\color{blue}76.41%}</math>]] | ||
− | |||
− | |||
| | | | ||
| | | | ||
− | | | + | | |
− | | | + | | |
+ | | | ||
+ | | | ||
| | | | ||
|- | |- | ||
Line 91: | Line 90: | ||
|We used [[Establishing an authenticity of sport news by Machine Learning Models#The RTextTools package|RTextTools]], which depends on [https://cran.r-project.org/web/packages/e1071/index.html e1071] | |We used [[Establishing an authenticity of sport news by Machine Learning Models#The RTextTools package|RTextTools]], which depends on [https://cran.r-project.org/web/packages/e1071/index.html e1071] | ||
|SVM* | |SVM* | ||
− | |||
|[[Establishing an authenticity of sport news by Machine Learning Models#fnd_svm|<math>{\color{blue}95.42%}</math>]] | |[[Establishing an authenticity of sport news by Machine Learning Models#fnd_svm|<math>{\color{blue}95.42%}</math>]] | ||
| | | | ||
Line 97: | Line 95: | ||
| | | | ||
| | | | ||
− | | | + | | |
| | | | ||
| | | | ||
Line 103: | Line 101: | ||
|[[Establishing an authenticity of sport news by Machine Learning Models#Random forest|Random forest]] | |[[Establishing an authenticity of sport news by Machine Learning Models#Random forest|Random forest]] | ||
|Liawand Wiener, 2002 | |Liawand Wiener, 2002 | ||
− | | | + | |[https://cran.r-project.org/web/packages/randomForest/index.html randomForest] |
|RF | |RF | ||
− | |||
| | | | ||
| | | | ||
Line 111: | Line 108: | ||
| | | | ||
| | | | ||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
| | | | ||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
| | | | ||
| | | | ||
Line 147: | Line 116: | ||
|[[Establishing an authenticity of sport news by Machine Learning Models#The XGBoost package|xgboost]] | |[[Establishing an authenticity of sport news by Machine Learning Models#The XGBoost package|xgboost]] | ||
|XGBOOST* | |XGBOOST* | ||
− | |||
|[[Establishing an authenticity of sport news by Machine Learning Models#fnd_xgboost|<math>{\color{red}97.42%}</math>]] | |[[Establishing an authenticity of sport news by Machine Learning Models#fnd_xgboost|<math>{\color{red}97.42%}</math>]] | ||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
| | | | ||
| | | | ||
Line 167: | Line 122: | ||
| | | | ||
| | | | ||
− | |||
| | | | ||
| | | | ||
|- | |- | ||
− | |[[Establishing an authenticity of sport news by Machine Learning Models# | + | |[[Establishing an authenticity of sport news by Machine Learning Models#General linearized models|General linearized models]] |
− | | | + | |Friedman et al., 2010 |
− | |We used [[Establishing an authenticity of sport news by Machine Learning Models#The RTextTools package|RTextTools]], which depends on [https://cran.r-project.org/web/packages/ | + | |We used [[Establishing an authenticity of sport news by Machine Learning Models#The RTextTools package|RTextTools]], which depends on [https://cran.r-project.org/web/packages/glmnet/index.html wglmnet] |
− | | | + | |GLMNET* |
− | |[[Establishing an authenticity of sport news by Machine Learning Models# | + | |[[Establishing an authenticity of sport news by Machine Learning Models#fnd_glmnet|<math>{\color{blue}94.58%}</math>]] |
| | | | ||
| | | | ||
Line 181: | Line 135: | ||
| | | | ||
| | | | ||
− | |||
| | | | ||
| | | | ||
|- | |- | ||
− | |[[Establishing an authenticity of sport news by Machine Learning Models# | + | |[[Establishing an authenticity of sport news by Machine Learning Models#Maximum entropy|Maximum entropy]] |
− | | | + | |Jurka, 2012 |
− | |We used [[Establishing an authenticity of sport news by Machine Learning Models#The RTextTools package|RTextTools]], which depends on | + | |We used [[Establishing an authenticity of sport news by Machine Learning Models#The RTextTools package|RTextTools]], which depends on [https://cran.r-project.org/web/packages/maxent/index.html maxent] |
− | [https://cran.r-project.org/web/packages/ | + | |MAXENT* |
− | | | + | |[[Establishing an authenticity of sport news by Machine Learning Models#fnd_maxent|<math>{\color{blue}96.09%}</math>]] |
− | |[[Establishing an authenticity of sport news by Machine Learning Models# | ||
| | | | ||
| | | | ||
Line 196: | Line 148: | ||
| | | | ||
| | | | ||
− | |||
| | | | ||
| | | | ||
|- | |- | ||
− | + | | colspan="12" |* Low-memory algorithm | |
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | |||
− | | colspan=" | ||
<nowiki>**</nowiki> Very high-memory algorithm | <nowiki>**</nowiki> Very high-memory algorithm | ||
+ | |||
+ | ^ | ||
+ | |||
+ | ^^ | ||
|} | |} | ||
Line 341: | Line 268: | ||
<br /> | <br /> | ||
− | |||
Revision as of 19:18, 27 April 2019
Contents
Declaration
Acknowledgement
Thanks for Muhammad, Graham and Mark
Abstract
Introduction
Chapter 1
Chapter 2 - Training a Supervised Machine Learning Model for fake news detection
Supervised text Classification for fake news detection Using Machine Learning Models
Procedure
- The Dataset
- Splitting the data into Train and Test data
- Cleaning the data
- Building the Document-Term Matrix
- Model Building
- Cross validation
- Making predictions from the model created and displaying a Confusion matrix
Results
Summary of Results
Algorithms | Author | Package | Keyword | Accuracy | |||||||
---|---|---|---|---|---|---|---|---|---|---|---|
Kaggle fake news dataset
20,800 rows |
Fake news Detector dataset
10,000 rows |
Gofaaas Fake News Dataset
500 rows | |||||||||
Test data (70% of the dataset) | Cross validation | Test data (70% of the dataset) | Cross validation | Test data (70% of the dataset) | Cross validation | Using the Kaggle Model^ | Using the Detector Model^^ | ||||
Naive Bayes | Bayes, Thomas | e1071 | NB* | ||||||||
Support vector machine | Meyer et al., 2012 | We used RTextTools, which depends on e1071 | SVM* | ||||||||
Random forest | Liawand Wiener, 2002 | randomForest | RF | ||||||||
Extreme Gradient Boosting | Chen & Guestrin, 2016 | xgboost | XGBOOST* | ||||||||
General linearized models | Friedman et al., 2010 | We used RTextTools, which depends on wglmnet | GLMNET* | ||||||||
Maximum entropy | Jurka, 2012 | We used RTextTools, which depends on maxent | MAXENT* | ||||||||
* Low-memory algorithm
** Very high-memory algorithm ^ ^^ |
Evaluation of Results
We evaluate our approach in different settings. First, weperform cross-validation on our noisy training set; second,and more importantly, we train models on the training setand validate them against a manually created gold standard.17Moreover, we evaluate two variants, i.e., including and exclud-ing user features. [smb:home/adelo/1-system/1-disco_local/1-mis_archivos/1-pe/1-ciencia/1-computacion/2-data_analysis-machine_learning/gofaaaz-machine_learning/5-References/7-Weakly_supervised_searning_for_fake_news_detection_on_twitter.pdf]
The Gofaaas-Fake News Detector R Package
Installation
Functions
Datasets used
Kaggle Fake News Dataset
https://www.kaggle.com/c/fake-news/data
Distribution of the data:
The distribution of Stance
classes in train_stances.csv
is as follows:
rows | unrelated | discuss | agree | disagree |
---|---|---|---|---|
49972 | 0.73131 | 0.17828 | 0.0736012 | 0.0168094 |
Fake News Detector Dataset
Gofaaas Fake News Dataset
Algorithms
Naive Bayes
Naïve Bayes is based on the Bayesian theorem, there in order to understand Naïve Bayes it is important to first understand the Bayesian theorem.
Bayesian theorem is a mathematical formula for determining conditional probability which is the probability of something, happening given that something else has already occurred.
- P(c|x) is the posterior probability of class (target) given predictor (attribute).
- P(c) is the prior probability of class.
- P(x|c) is the likelihood which is the probability of predictor given class.
- P(x) is the prior probability of predictor.
Prior probability, in Bayesian statistical inference, is the probability of an event before new data is collected.
Posterior probability is the revised probability of an event occurring after taking into consideration new information.
In statistical terms, the posterior probability is the probability of event A occurring given that event B has occurred.
Support vector machine
Random forest
Extreme Gradient Boosting
The XGBoost R package
XGBoost - Extreme Gradient Boosting:
- https://xgboost.readthedocs.io/en/latest/
- https://cran.r-project.org/web/packages/xgboost/index.html
The RTextTools package
RTextTools - A Supervised Learning Package for Text Classification:
- https://journal.r-project.org/archive/2013/RJ-2013-001/RJ-2013-001.pdf
- http://www.rtexttools.com/
- https://cran.r-project.org/web/packages/RTextTools/index.html
Chapter 3 - Gofaas Web App
A way to interact, test and display the model results
Conclusion