Corporate governance and IFRS in Saudi ArabiaThe best way for Data evaluation

1

Nouh, M., Nurse, J. R., & Goldsmith, M. (2019). POSTER: Detection of Online Radical Content Using Multimodal Approach. Detection of Online Radical Content, 01(10), 01-10.

2019

POSTER: Detection of Online Radical Content Using Multimodal Approach

The research was focused on studying the online behavior of the users and on performing the content-based analysis that works to distinguish textual features. In the research proposed research was focused on radicalized content and the challenging task.

The subject of the research was to detect the online radical content. 

In this study, the techniques that were used is “Machine learning methodology” which classify tweets along with the identification of the approach. And the unsupervised “Machine learning method “is also used where cluster online texts the messages into the group radical messages.

One of the limitations of the research was that the selected approach was highly based on the data. The issue was related to domain selection and identification of terrorist networks.

The approach can be used to detect radicalized online messages. The effort is similar to the lexical analysis, but it can be used in the future for the detailed analysis of the Twitter platform. Pro-Extremist users  can be used to identify the patterns that predict psychological probability.

2

Neethu, S., & Rajasree, R. (2013). Sentiment Analysis in Twitter using Machine Learning Techniques. 4th ICCCNT 2013, 04 (06), 01-10

2013

Sentiment Analysis in Twitter using Machine Learning Techniques

The aim of the research is to use symbolic techniques based on the knowledge approach and machine learning techniques. Sentiment analysis is also used to determine the levels of topics on the Twitter platform. The feature vector performance is also evaluated for the electronic product domains.

There are some specific feature extraction methods that can be used to extract relevant information. In the process, the standard database is not available to analyze the Twitter sentiment and the standard database.

There are different methodologies were used in this study to detect the sentiments by considering the “machine learning techniques, SVM, naïve Bayes. In the research, automatic sentiment analysis was used to classify the tweets and for the normalization of knowledge base approach. And to estimates the probability of eth positivity tweets, then this study were used the SVM, Naïve Bayes techniques.  

The sentimental analysis deal with slang words and misspelling. The research was limited for a number of characters as 140 characters in both knowledge base approach and machine learning approach.

It can be used to deal with the feature extraction process, efficient feature vector, and proper processing of some specific features. The proper classification accuracy can be used to test different types of classifiers such as maximum entropy, SVM, Nave Bayes and electronic product domains.

3

Caplan, J. (2013). Social Media and Politics: Twitter Use in the Second Congressional District of Virginia. Elon Journal of Undergraduate Research in Communications, 04 (01), 01-06.

2013

Social Media and Politics: Twitter Use in the

Second Congressional District of Virginia

The goal of the study was to get a proper insight into the process in which how Republican Congressman Scott Rigell and Democratic

candidate Paul Hirschbiel—candidates in the 2nd Congressional District of Virginia use Twitter to reveal the post information. In this process, direct communication tweets were used to post information and to use tactical strategies for the determination of calculated methods to motivate the citizens.

The aim was to measure evaluation in the informal, political exchange and the social influence on the targeted numerous sectors of American society. The research questions were designed on the basis of propelled studies, characteristic, and tactical strategies with high intensity. On the basis of this discussion, some content related to the incumbent was analyzed for social media of Scott Rigell and Paul hirschbiel. In the process, the election was conducted on Twitter.

The data gathering and the coding technique were used in this research study. In the research, Hampton Roads, the 2nd Congressional District of Virginia, was highly saturated with the political advertainment before the congressional election. The study analyzed the Twitter strategies of Republican incumbent Scott Rigell and the competitive Paul Hirschbiel. The other competition was also measured for the candidates to differentiate them from the active voters.  

The research is limited to analyze the use of Twitter in the election campaign of Republican Congressman Scott Rigell and Democratic

candidate Paul Hirschbiel—candidates in the 2nd Congressional District of Virginia. The research is limited to direct contact with the potential supporters and testimonials of platforms for the involvement of the community. It was analyzed that how a community is involved with social media for the election. The potential impact of Twitter is measured on the election. 

In future, the study can be used for the analysis of responses of the Twitter users towards the election and researcher can collect data from these responses. The candidate distinctively uses Twitter to attract the voters. 

4

Mittal, A., & Goel, A. (2012). Stock Prediction Using Twitter Sentiment Analysis. Standford University, CS229, 15 (01), 01-10.

2012

Stock Prediction Using Twitter Sentiment Analysis.

The research used sentiment analysis and the machine learning process. In the research, the correlation was determined to use the twitter data and prediction of the public mood. The results were testified by the cross-validation method related to the financial data. Some percentage of accuracy related to the neural network can be used. The management strategy was used to predict values along with accuracy.

In the research paper, the hypothesis was based on the premise of behavioural economics, and it analyzed the impact of moods and emotions on the decision-making process. the research determined the correlation between the market sentiment and the public sentiment. The degree of membership and the data available at twitter was used in the analysis. 

In this research study, the techniques were used is that, Linear regression, Logistics regression. SVMs, these techniques are sued after finding the causality relationships among the 3 past days of moods plus currents of the day stock prices; then tried these techniques.  This research is used the  LIBSVM for the SVM techniques.

The research was limited for two parameters for the optimal parametrization. In the research, large enough test sets were used, but limited data was used with 30 to 40 entries.

SOFNN model can be used to predict the future values of DJIA algorithms. The model can predict future values.

5

Benhardus, J. (2013). Streaming trend detection in Twitter. Int. J. Web Based Communities, 09 (01), 122-130.

2013

Streaming trend detection in Twitter

The research provided with the traditional media identification process. the trending process was based on the issues and overviews. In the research outline methods were used to detect and identify the trending topics for the data streaming.  The research used data from Twitter streaming and then analyzed trending topics on Twitter.

In the research, the term frequency-inverse document was used to identify trending topics. The results of the methodologies after the streaming data was measured by having the F-measure range at different human language processes.  

The techniques were used in this study are NPL, to streaming the data from Twitter, which identifies the trending topics.   Different procedures were used for the data collection and allowed analysis through the multiple timespans and trending topics of twitter. The term frequency-inverse document frequency analysis and the relative normalized term frequency analysis were used in the research to measure the trending topics. The relative normalized term frequency used trigrams, bigrams, and unigrams.

The limited stream data was used for the garden hose streaming, and Twitter activity was limited to 15% only. The research is only limited to some models and methods such as normalization of term frequency analysis, identification of trending topics on Twitter and reasonable analysis for the success conditions of Twitter users. The possible extensions are limited for maximum improvement, and it can be used only for the limited topics to be used in the bigram and unigram models.

Even the model achieved success, but there are some possibilities to improve the extension and to use different methods for the detection of potential extension.

6

Kouloumpis, E., Wilson, T., & Moore, J. (2011). Twitter Sentiment Analysis: The Good the Bad and the OMG! Proceedings of the Fifth International AAAI Conference on Weblogs and Social Media, 01 (10), 538-542

2011

Twitter Sentiment Analysis: The Good the Bad and the OMG!

The research identified the linguistic features for the determination and detection of the sentiment of messages on Twitter. The other consideration of the work was to measure the existing lexical resources with certain features. Some superficial approaches were used for training data analysis.

The main objective of the research was to identify the microblogging platform of Twitter for the services and products. The fair amount of research was expected for the incredible breadth of the topic. In the research, quick identification was used for building data such as news or best feelings.

The technique was used in this study is twitter hashtags, like #epicfail, #news, #bestfeelings. These techniques identified the positive and negative, along with neutral tweets to use the three methods of sentiments classifiers.   The twitter sentiment analysis was used in different ways for the unique featuring techniques. The researchers investigated the automatic process for the level classification. The researchers investigated various methods for automatic data collection and sentimental classification.

The research is limited to some specific consideration related to the experiments, sentiments and different sentimental techniques. The limited data was used for the 3-way polarity of classification.

The research can be extended further by collecting data based on positive and negative emotions. However, there are two resources related to the methods used in the features and complementary process. the microblogging features can be used in the experiment related to benefits of the emotion and training data.

7

SUN, A., NAING, M.-M., LIM, E. P., & LAM, W. (2003). Using Support Vector Machines for Terrorism Information Extraction. Research Collection School of Information Systems, 26(06), 01-12

2003

Using Support Vector Machines for Terrorism Information Extraction.

Use of information extraction in the research is significant for web intelligence and the IE systems. The extracted pattern was different fort he considered constraint and the semantic position with the wider range. The research considered extraction patterns with restricted templates. The process of modelling was linked with the context of information.

In the research, the classification model was used to develop support for the vector machine. The IE experiments were used to evaluate the proposed methods for the text collection and the terrorism domain. The preliminary experimental work was used to evaluate the proposed methods and to deliver reasonable accuracies. 

The techniques were used in this study are IE, and the SVM. The research used the IE method for SVM techniques to extract the data. The IE support was used for the support vector machines SVM, and the successful transformation was used to extract the target entities.The template element extraction task was used to refer to the semantic categories. The documents were used for the terrorism domains, including perpetrator, victim and witness.

 The research is limited to the extraction of perpetrator entities with the collection of untagged documents. The SVM classification was also used.

The exploratory research generalized the data and some of the featured selection can be considered for the performance evaluation in future research.