The 21st century has been characterized by an increased attention to social networks. Nowadays, going 24 hours without getting in touch with them in some way has become difficult. Facebook and Twitter, these social platforms are now part of everyday life. Thus, these social networks have become important sources to be aware of frequently discussed topics or public opinions on a current issue. A lot of people write messages about current events, give their opinion on any topic and discuss social issues more and more.
The impact of sentiment analysis from user on Facebook to enhanced the servic...IJECEIAES
Facebook's influence on the modern social media platform is undoubtedly enormous. While it has gotten a backlash for its inability to control its influence over important affairs, there are still many questions regarding people's perception of Facebook and their sentiment over Facebook. This paper's role in this ongoing debate is to give a glimpse of people's sentiment and perception of Facebook in recent times. By collecting samples data from Facebook's Top Page, this paper hopes to represent a significant amount of people's aspirations towards this company. By processing the data with a processing tool to construct and model out the data and a sentiment analyzer tool helps determine the sentiment, this paper can deduce a 600-comment worth of processed data. The results from the 600 sampled comments concluded that the sentiments towards Facebook are 41.50% negative comments, 22.83% neutral comments, and 35.67% positive comments.
The development of new technologies will enable decentralization and freedom of communication for large numbers of people, by overcoming the barriers that once rendered direct participation of
society unfeasible. The continued development of information and communication technologies (ICT)
makes it possible for people to participate in political life. Today, the use of e-tools is becoming a way of
adapting democracy to the needs of contemporary states and strengthening civil society. The aim of this
paper is to answer questions about the essence of ICT and forms of civic engagement through electronic
forms of participation. The author seeks answers to the following questions: How does ICT influence
political processes? How do electronic communication systems create the conditions for the political engagement of citizens? Can the use of information technologies have a real impact on participation?
The impact of sentiment analysis from user on Facebook to enhanced the servic...IJECEIAES
Facebook's influence on the modern social media platform is undoubtedly enormous. While it has gotten a backlash for its inability to control its influence over important affairs, there are still many questions regarding people's perception of Facebook and their sentiment over Facebook. This paper's role in this ongoing debate is to give a glimpse of people's sentiment and perception of Facebook in recent times. By collecting samples data from Facebook's Top Page, this paper hopes to represent a significant amount of people's aspirations towards this company. By processing the data with a processing tool to construct and model out the data and a sentiment analyzer tool helps determine the sentiment, this paper can deduce a 600-comment worth of processed data. The results from the 600 sampled comments concluded that the sentiments towards Facebook are 41.50% negative comments, 22.83% neutral comments, and 35.67% positive comments.
The development of new technologies will enable decentralization and freedom of communication for large numbers of people, by overcoming the barriers that once rendered direct participation of
society unfeasible. The continued development of information and communication technologies (ICT)
makes it possible for people to participate in political life. Today, the use of e-tools is becoming a way of
adapting democracy to the needs of contemporary states and strengthening civil society. The aim of this
paper is to answer questions about the essence of ICT and forms of civic engagement through electronic
forms of participation. The author seeks answers to the following questions: How does ICT influence
political processes? How do electronic communication systems create the conditions for the political engagement of citizens? Can the use of information technologies have a real impact on participation?
Chung-Jui LAI - Polarization of Political Opinion by News MediaREVULN
In 2016 US election, social media played a vital role in shaping public opinions as expressed by the news media that have created the phenomenon of polarization in the United States. Because social media gave people the ability to follow, share, post, comment below everything, the phenomenon of political opinions being spread easily and quickly on social media by the news agencies is bringing out a significantly polarized populace.
Consequently, it’s very important to understand the language differences on Twitter and figure out how propaganda spread by different political parties that influence or perhaps mislead public opinion. This talk will introduce the relationship among the social media, public opinion, and news media, then suggests the method to collect the tweets from Twitter and conduct sentimental and logistic regression analysis on them. Furthermore, this talk points out the special aspect on the relationship between the polarization and the topic of this conference (fake news, disinformation and propaganda).
Main points:
- situation in Taiwan
- research on fake news
- methods for fighting fake news
How does fakenews spread understanding pathways of disinformation spread thro...Araz Taeihagh
What are the pathways for spreading disinformation on social media platforms? This article addresses this question by collecting, categorising, and situating an extensive body of research on how application programming interfaces (APIs) provided by social media platforms facilitate the spread of disinformation. We first examine the landscape of official social media APIs, then perform quantitative research on the open-source code repositories GitHub and GitLab to understand the usage patterns of these APIs. By inspecting the code repositories, we classify developers' usage of the APIs as official and unofficial, and further develop a four-stage framework characterising pathways for spreading disinformation on social media platforms. We further highlight how the stages in the framework were activated during the 2016 US Presidential Elections, before providing policy recommendations for issues relating to access to APIs, algorithmic content, advertisements, and suggest rapid response to coordinate campaigns, development of collaborative, and participatory approaches as well as government stewardship in the regulation of social media platforms.
E-consultations: New tools for civic engagement or facades for political corr...ePractice.eu
Author: Jordanka Tomkova.
Since the 1990s, public institutions have been increasingly reaching into democracy's toolbox for new tools with which to better engage citizens in politics.
V Międzynarodowa Konferencja Naukowa Nauka o informacji (informacja naukowa) w okresie zmian Innowacyjne usługi informacyjne. Wydział Dziennikarstwa, Informacji i Bibliologii Katedra Informatologii, Uniwersytet Warszawski, Warszawa, 15 – 16 maja 2017
Fredrick Ishengoma - Online Social Networks and Terrorism 2.0 in Developing C...Fredrick Ishengoma
The advancement in technology has brought a new era in terrorism where Online Social Networks (OSNs) have become a major platform of communication with wide range of usage from message channeling to propaganda and recruitment of new followers in terrorist groups. Meanwhile, during the terrorist attacks people use OSNs for information exchange, mobilizing and uniting and raising money for the victims. This paper critically analyses the specific usage of OSNs in the times of terrorisms attacks in developing countries. We crawled and used Twitter’s data during Westgate shopping mall terrorist attack in Nairobi, Kenya. We then analyzed the number of tweets, geo-location of tweets, demographics of the users and whether users in developing countries tend to tweet, retweet or reply during the event of a terrorist attack. We define new metrics (reach and impression of the tweet) and present the models for calculating them. The study findings show that, users from developing countries tend to tweet more at the first and critical times of the terrorist occurrence. Moreover, large number of tweets originated from the attacked country (Kenya) with 73% from men and 23% from women where original posts had a most number of tweets followed by replies and retweets.
Who’s in the Gang? Revealing Coordinating Communities in Social MediaDerek Weber
Political astroturfing and organised trolling are online malicious behaviours with significant real-world effects. Common approaches examining these phenomena focus on broad campaigns rather than the small groups responsible. To reveal networks of cooperating accounts, we propose a novel temporal window approach that relies on account interactions and metadata alone. It detects groups of accounts engaging in behaviours that, in concert, execute different goal-based strategies, which we describe. Our approach is validated against two relevant datasets with ground truth data. See https://github.com/weberdc/find_hccs for code and data.
Presented at ASONAM'20 (2020 IEEE/ACM International Conference on Advances in Social Network Analysis and Mining).
Co-authored with Frank Neumann (University of Adelaide)
This paper examines digital literacy and how it relates to the philosophical study of ignorance. Ignorance of how digital technologies work (e.g. how users’ online activities can be used to the advantage of platform owners without the users’ knowledge, and how browsing can be confined) is still not well understood from the perspective of user practice.
Based on the following Special Issue of Teaching in Higher Education: https://doi.org/10.1080/13562517.2018.1547276
Talk done at Lancaster University, Edinburgh University, the SRHE conference, Sussex University,
This paper aims to examine how political conversations take place on the digital
discursive tools offered as part of the Digital Participatory Budget (OPD) in Belo Horizonte (Brazil). The authors propose an analytical model based on deliberative theories in order to investigate the discussions over this participatory program. The main sample consists of the messages posted by the users (n=375) on the commentaries section. The results show that reciprocity and reflexivity among interlocutors are rare; however, the respect among the participants and the justification levels in several arguments were high during the discussion. The authors conclude that, even in a
situation in which there is no empowerment of the digital tools, the internet can effectively provide environments to enhance a qualified discursive exchange. In spite of low levels of deliberativeness, the case study shows that there are important gains concerning social learning among the participants.
This collaborative research-project between Global Pulse (www.unglobalpulse.org) and SAS (www.sas.com) investigates how social media and online user-generated content can be used to enrich the understanding of the changing job conditions in the US and Ireland by analyzing the moods and topics present in unemployment-related conversations from the open social web and relating them to official unemployment statistics. For more information on this project or the other projects in this series, please visit: http://www.unglobalpulse.org/research.
Under what conditions can information and communications technologies (ICTs) enhance the well-being of poor communities? The paper designs an alternative evaluation framework (AEF) that applies Sen’s capability approach to the study of ICTs in order to place people’s well-being, rather than technology at the center of the study. The AEF develops an impact chain that examines the mechanisms by which access to, and meaningful use of, ICTs can enhance peoples “informational capabilities” and can lead to improvements in people’s human and social capabilities. This approach thus uses peoples’ human capabilities, rather than measures of access or usage, as its principal evaluative space. Based on empirical evidence from rural communities’ uses of ICTs in Bolivia, the study concludes that enhancing people’s informational capabilities is the most critical factor determining the impact of ICTs on their well-being. The findings indicate that improved informational capabilities, like literacy, do enhance the human capabilities of the poor and marginalized to make strategic life choices to achieve the lifestyle they value. Evaluating the impact of ICTs in terms of capabilities thus reveals that there is no direct relationship between improved access to, and use of, ICTs and enhanced well-being; ICTs lead to improvements in people’s lives only when informational capabilities are transformed into expanded human and social capabilities in the economic, political, social, organizational and cultural dimensions of their lives.
Amplification and Personalization: The impact of metrics, analytics, and algo...Nicole Blanchett
This presentation was done for the Stratford chapter of CFUW and focuses on the impact of metrics and analytics on information sharing, how what you consume on the web leaves a trail of behaviour patterns that allow for personalization of content, and how everyone can help stop the amplification of misinformation.
What Data Can Do: A Typology of Mechanisms
Angèle Christin .
International Journal of Communication > Vol 14 (2020) , de Angèle Christin del Departamento de Comunicación de Stanford University, USA titulado "What Data Can Do: A Typology of Mechanisms". Entre otras cosas es autora del libro "Metrics at Work.
The Pessimistic Investor Sentiments Indicator in Social NetworksTELKOMNIKA JOURNAL
With the worldwide proliferation of social networks, the social networks have played an important role in the social activities .Peoples are inclined to obtain the corresponding public opinion to make decision such as shopping, education, investment and so on. Analysis of data generated by social networks has become an important field of research, however in the field of public opinion analysis of social networks the quantitative measure indexes are still lacking. In this paper, the calculation method of pessimistic investor sentiments indicator is proposed, and the index has a certain theoretical and practical value.
Need Response 1The subcomponent of crowdsourcing ICT platform.docxvannagoforth
Need Response 1:
The subcomponent of crowdsourcing ICT platform technological architecture I would like to discuss is that gives additionally created an examination concerning public text information (for instance blog postings, comments, appraisals, etc.) and setting up this information sources using sentiment mining tools. The Web has changed the way wherein people express their emotions, offering them the capacity to post comments and reviews on business things and express their points of view on a a huge amount of issues in parties, talk get-togethers, visit rooms, long-broaden agreeable correspondence get-togethers and web diaries. This customer passed on the substance has been seen as a tremendous wellspring of business and political information. Notwithstanding, the tremendous the degree of this information and its normal language structure makes it difficult to remove the consistent areas, for instance, the general inclination/assessment (for instance positive, negative or sensible) on the particular subject (for instance a thing/affiliation or another methodology proposition) and the specific issues raised about it by the customers/visitors of these objectives. It is hidden motivation has been to enable firms to research online overviews and comments entered by customers of their things in various review districts, web diaries, social affairs, etc., in order to arrive at general judgments as for whether customers adored the thing or not (supposition assessment), and moreover continuously express finishes concerning features (traits) of the thing that has been commented on insistently or conflictingly (features extraction and examination).
This subcomponent performs three tasks, firstly it classifies the opinion text, a document which includes various declarations like a dialogue or a blog spot conveying a positive, negative or unprejudiced end. This is suggested as the record level evaluation examination. Secondly further focusing on sentence-level which deals with the gathering of a sentence as objective or passionate, it organizes each sentence in such a structure, that atmosphere it is a unique or targets (demonstrating whether it can express the inclination or not). For each sentence that is a conceptual (infers conveying an inclination) further, the portrayal is done as imparting an appositive, negative or unprejudiced supposition. Lastly extracting the most commented features of the commented articles, and for each commented feature further classification of relevant opinion is executed as positive, negative or unprejudiced.
References
Janssen, M., Wimmer, M. A., & Deljoo, A. (Eds.). (2015). Policy practice and digital science: Integrating complex systems, social simulation and public administration in policy research (Vol. 10). Springer
Need response 2:
nformation and communication technology platform has an important role to play in active crowdsourcing. A policy maker of a government agency initiates ...
Chung-Jui LAI - Polarization of Political Opinion by News MediaREVULN
In 2016 US election, social media played a vital role in shaping public opinions as expressed by the news media that have created the phenomenon of polarization in the United States. Because social media gave people the ability to follow, share, post, comment below everything, the phenomenon of political opinions being spread easily and quickly on social media by the news agencies is bringing out a significantly polarized populace.
Consequently, it’s very important to understand the language differences on Twitter and figure out how propaganda spread by different political parties that influence or perhaps mislead public opinion. This talk will introduce the relationship among the social media, public opinion, and news media, then suggests the method to collect the tweets from Twitter and conduct sentimental and logistic regression analysis on them. Furthermore, this talk points out the special aspect on the relationship between the polarization and the topic of this conference (fake news, disinformation and propaganda).
Main points:
- situation in Taiwan
- research on fake news
- methods for fighting fake news
How does fakenews spread understanding pathways of disinformation spread thro...Araz Taeihagh
What are the pathways for spreading disinformation on social media platforms? This article addresses this question by collecting, categorising, and situating an extensive body of research on how application programming interfaces (APIs) provided by social media platforms facilitate the spread of disinformation. We first examine the landscape of official social media APIs, then perform quantitative research on the open-source code repositories GitHub and GitLab to understand the usage patterns of these APIs. By inspecting the code repositories, we classify developers' usage of the APIs as official and unofficial, and further develop a four-stage framework characterising pathways for spreading disinformation on social media platforms. We further highlight how the stages in the framework were activated during the 2016 US Presidential Elections, before providing policy recommendations for issues relating to access to APIs, algorithmic content, advertisements, and suggest rapid response to coordinate campaigns, development of collaborative, and participatory approaches as well as government stewardship in the regulation of social media platforms.
E-consultations: New tools for civic engagement or facades for political corr...ePractice.eu
Author: Jordanka Tomkova.
Since the 1990s, public institutions have been increasingly reaching into democracy's toolbox for new tools with which to better engage citizens in politics.
V Międzynarodowa Konferencja Naukowa Nauka o informacji (informacja naukowa) w okresie zmian Innowacyjne usługi informacyjne. Wydział Dziennikarstwa, Informacji i Bibliologii Katedra Informatologii, Uniwersytet Warszawski, Warszawa, 15 – 16 maja 2017
Fredrick Ishengoma - Online Social Networks and Terrorism 2.0 in Developing C...Fredrick Ishengoma
The advancement in technology has brought a new era in terrorism where Online Social Networks (OSNs) have become a major platform of communication with wide range of usage from message channeling to propaganda and recruitment of new followers in terrorist groups. Meanwhile, during the terrorist attacks people use OSNs for information exchange, mobilizing and uniting and raising money for the victims. This paper critically analyses the specific usage of OSNs in the times of terrorisms attacks in developing countries. We crawled and used Twitter’s data during Westgate shopping mall terrorist attack in Nairobi, Kenya. We then analyzed the number of tweets, geo-location of tweets, demographics of the users and whether users in developing countries tend to tweet, retweet or reply during the event of a terrorist attack. We define new metrics (reach and impression of the tweet) and present the models for calculating them. The study findings show that, users from developing countries tend to tweet more at the first and critical times of the terrorist occurrence. Moreover, large number of tweets originated from the attacked country (Kenya) with 73% from men and 23% from women where original posts had a most number of tweets followed by replies and retweets.
Who’s in the Gang? Revealing Coordinating Communities in Social MediaDerek Weber
Political astroturfing and organised trolling are online malicious behaviours with significant real-world effects. Common approaches examining these phenomena focus on broad campaigns rather than the small groups responsible. To reveal networks of cooperating accounts, we propose a novel temporal window approach that relies on account interactions and metadata alone. It detects groups of accounts engaging in behaviours that, in concert, execute different goal-based strategies, which we describe. Our approach is validated against two relevant datasets with ground truth data. See https://github.com/weberdc/find_hccs for code and data.
Presented at ASONAM'20 (2020 IEEE/ACM International Conference on Advances in Social Network Analysis and Mining).
Co-authored with Frank Neumann (University of Adelaide)
This paper examines digital literacy and how it relates to the philosophical study of ignorance. Ignorance of how digital technologies work (e.g. how users’ online activities can be used to the advantage of platform owners without the users’ knowledge, and how browsing can be confined) is still not well understood from the perspective of user practice.
Based on the following Special Issue of Teaching in Higher Education: https://doi.org/10.1080/13562517.2018.1547276
Talk done at Lancaster University, Edinburgh University, the SRHE conference, Sussex University,
This paper aims to examine how political conversations take place on the digital
discursive tools offered as part of the Digital Participatory Budget (OPD) in Belo Horizonte (Brazil). The authors propose an analytical model based on deliberative theories in order to investigate the discussions over this participatory program. The main sample consists of the messages posted by the users (n=375) on the commentaries section. The results show that reciprocity and reflexivity among interlocutors are rare; however, the respect among the participants and the justification levels in several arguments were high during the discussion. The authors conclude that, even in a
situation in which there is no empowerment of the digital tools, the internet can effectively provide environments to enhance a qualified discursive exchange. In spite of low levels of deliberativeness, the case study shows that there are important gains concerning social learning among the participants.
This collaborative research-project between Global Pulse (www.unglobalpulse.org) and SAS (www.sas.com) investigates how social media and online user-generated content can be used to enrich the understanding of the changing job conditions in the US and Ireland by analyzing the moods and topics present in unemployment-related conversations from the open social web and relating them to official unemployment statistics. For more information on this project or the other projects in this series, please visit: http://www.unglobalpulse.org/research.
Under what conditions can information and communications technologies (ICTs) enhance the well-being of poor communities? The paper designs an alternative evaluation framework (AEF) that applies Sen’s capability approach to the study of ICTs in order to place people’s well-being, rather than technology at the center of the study. The AEF develops an impact chain that examines the mechanisms by which access to, and meaningful use of, ICTs can enhance peoples “informational capabilities” and can lead to improvements in people’s human and social capabilities. This approach thus uses peoples’ human capabilities, rather than measures of access or usage, as its principal evaluative space. Based on empirical evidence from rural communities’ uses of ICTs in Bolivia, the study concludes that enhancing people’s informational capabilities is the most critical factor determining the impact of ICTs on their well-being. The findings indicate that improved informational capabilities, like literacy, do enhance the human capabilities of the poor and marginalized to make strategic life choices to achieve the lifestyle they value. Evaluating the impact of ICTs in terms of capabilities thus reveals that there is no direct relationship between improved access to, and use of, ICTs and enhanced well-being; ICTs lead to improvements in people’s lives only when informational capabilities are transformed into expanded human and social capabilities in the economic, political, social, organizational and cultural dimensions of their lives.
Amplification and Personalization: The impact of metrics, analytics, and algo...Nicole Blanchett
This presentation was done for the Stratford chapter of CFUW and focuses on the impact of metrics and analytics on information sharing, how what you consume on the web leaves a trail of behaviour patterns that allow for personalization of content, and how everyone can help stop the amplification of misinformation.
What Data Can Do: A Typology of Mechanisms
Angèle Christin .
International Journal of Communication > Vol 14 (2020) , de Angèle Christin del Departamento de Comunicación de Stanford University, USA titulado "What Data Can Do: A Typology of Mechanisms". Entre otras cosas es autora del libro "Metrics at Work.
The Pessimistic Investor Sentiments Indicator in Social NetworksTELKOMNIKA JOURNAL
With the worldwide proliferation of social networks, the social networks have played an important role in the social activities .Peoples are inclined to obtain the corresponding public opinion to make decision such as shopping, education, investment and so on. Analysis of data generated by social networks has become an important field of research, however in the field of public opinion analysis of social networks the quantitative measure indexes are still lacking. In this paper, the calculation method of pessimistic investor sentiments indicator is proposed, and the index has a certain theoretical and practical value.
Need Response 1The subcomponent of crowdsourcing ICT platform.docxvannagoforth
Need Response 1:
The subcomponent of crowdsourcing ICT platform technological architecture I would like to discuss is that gives additionally created an examination concerning public text information (for instance blog postings, comments, appraisals, etc.) and setting up this information sources using sentiment mining tools. The Web has changed the way wherein people express their emotions, offering them the capacity to post comments and reviews on business things and express their points of view on a a huge amount of issues in parties, talk get-togethers, visit rooms, long-broaden agreeable correspondence get-togethers and web diaries. This customer passed on the substance has been seen as a tremendous wellspring of business and political information. Notwithstanding, the tremendous the degree of this information and its normal language structure makes it difficult to remove the consistent areas, for instance, the general inclination/assessment (for instance positive, negative or sensible) on the particular subject (for instance a thing/affiliation or another methodology proposition) and the specific issues raised about it by the customers/visitors of these objectives. It is hidden motivation has been to enable firms to research online overviews and comments entered by customers of their things in various review districts, web diaries, social affairs, etc., in order to arrive at general judgments as for whether customers adored the thing or not (supposition assessment), and moreover continuously express finishes concerning features (traits) of the thing that has been commented on insistently or conflictingly (features extraction and examination).
This subcomponent performs three tasks, firstly it classifies the opinion text, a document which includes various declarations like a dialogue or a blog spot conveying a positive, negative or unprejudiced end. This is suggested as the record level evaluation examination. Secondly further focusing on sentence-level which deals with the gathering of a sentence as objective or passionate, it organizes each sentence in such a structure, that atmosphere it is a unique or targets (demonstrating whether it can express the inclination or not). For each sentence that is a conceptual (infers conveying an inclination) further, the portrayal is done as imparting an appositive, negative or unprejudiced supposition. Lastly extracting the most commented features of the commented articles, and for each commented feature further classification of relevant opinion is executed as positive, negative or unprejudiced.
References
Janssen, M., Wimmer, M. A., & Deljoo, A. (Eds.). (2015). Policy practice and digital science: Integrating complex systems, social simulation and public administration in policy research (Vol. 10). Springer
Need response 2:
nformation and communication technology platform has an important role to play in active crowdsourcing. A policy maker of a government agency initiates ...
PREDICTING ELECTION OUTCOME FROM SOCIAL MEDIA DATAkevig
In this era of technology, enormous Online Social Networking Sites (OSNs) have arisen as a medium of expressing any opinions, thoughts towards anything even support their status against any social or political matter at the same time. Nowadays, people connected to those networks are more likely to prefer to employ themselves utilizing these online platforms to exhibit their standings upon any political organizations
participating in the election throughout the whole election period. The aim of this paper is to predict the outcome of the election by engaging the tweets posted on Twitter pertaining to the Australian federal election-2019 held on May 18, 2019. We aggregated two efficacious techniques in order to extract the
information from the tweet data to count a virtual vote for each corresponding political group. The original results of the election closely match the findings of our investigation, published by the Australian Electoral Commission.
PREDICTING ELECTION OUTCOME FROM SOCIAL MEDIA DATAijnlc
In this era of technology, enormous Online Social Networking Sites (OSNs) have arisen as a medium of expressing any opinions, thoughts towards anything even support their status against any social or political matter at the same time. Nowadays, people connected to those networks are more likely to prefer to employ themselves utilizing these online platforms to exhibit their standings upon any political organizations participating in the election throughout the whole election period. The aim of this paper is to predict the outcome of the election by engaging the tweets posted on Twitter pertaining to the Australian federal election-2019 held on May 18, 2019. We aggregated two efficacious techniques in order to extract the information from the tweet data to count a virtual vote for each corresponding political group. The original results of the election closely match the findings of our investigation, published by the Australian Electoral Commission.
PREDICTING ELECTION OUTCOME FROM SOCIAL MEDIA DATAkevig
In this era of technology, enormous Online Social Networking Sites (OSNs) have arisen as a medium of
expressing any opinions, thoughts towards anything even support their status against any social or
political matter at the same time. Nowadays, people connected to those networks are more likely to prefer
to employ themselves utilizing these online platforms to exhibit their standings upon any political
organizations participating in the election throughout the whole election period. The aim of this paper is to
predict the outcome of the election by engaging the tweets posted on Twitter pertaining to the Australian
federal election-2019 held on May 18, 2019. We aggregated two efficacious techniques in order to extract
the information from the tweet data to count a virtual vote for each corresponding political group. The
original results of the election closely match the findings of our investigation, published by the Australian
Electoral Commission.
PREDICTING ELECTION OUTCOME FROM SOCIAL MEDIA DATAkevig
In this era of technology, enormous Online Social Networking Sites (OSNs) have arisen as a medium of
expressing any opinions, thoughts towards anything even support their status against any social or
political matter at the same time. Nowadays, people connected to those networks are more likely to prefer
to employ themselves utilizing these online platforms to exhibit their standings upon any political
organizations participating in the election throughout the whole election period. The aim of this paper is to
predict the outcome of the election by engaging the tweets posted on Twitter pertaining to the Australian
federal election-2019 held on May 18, 2019. We aggregated two efficacious techniques in order to extract
the information from the tweet data to count a virtual vote for each corresponding political group. The
original results of the election closely match the findings of our investigation, published by the Australian
Electoral Commission.
Twitter is going to become much less important as a person to person (peer to peer) communication medium and instead become more of a content-delivery medium like TV, where content is broadcast to a large number of followers. It's just going to become a new way to follow celebrities, corporations, and the like.
These are the findings of research by two business school professors. Their paper "Intrinsic versus Image-Related Motivations in Social Media: Why do People Contribute to Twitter?", was published in the journal Marketing Science. The professors are Olivier Toubia of Columbia Business School and Andrew T. Stephen of the University of Pittsburgh.
In Professor Toubia’s words, “Get ready for a TV-like Twitter".
The research examined the motivations behind why everyday people, with no financial incentive, contribute to Twitter.The study examined roughly 2500 non-commercial Twitter users. In a field experiment, the profs randomly selected some of those users and, through the use of other synthetic accounts, increased the selected group's followers. At first, they noticed that as the selected group's followers increased, so did the posting rate. However, when that group reached a level of stature — a moderately large amount of followers — the posting rate declined significantly.
"Users began to realize it was harder to continue to attract more followers with their current strategy, so they slowed down.When posting activity no longer leads to additional followers, people will view Twitter as a non-evolving, static structure, like TV."
Based on the analyses, the profs predict Twitter posts by everyday people will slow down, yet celebrities and commercial users will continue to post for financial gain.
The paper is attached on Slideshare.
Running head: ANNOTATED OUTLINE 1
ANNOTATED OUTLINE 5
Annotated Outline
“The Impact of Social Media on Opinion of People”
Student Name
Course Name
Professor Name
Institution Name
Date:
Annotated Outline
Title:
“The Impact of Social Media on Opinion of People”
Introduction:
Social Media is a new place for people to learn, teach, and get knowledge or information about anything. This paper focuses on the importance of social media and its impact on the opinion of the people. More than half of the world is now using different social media platforms such as Facebook, Twitter, Instagram, etc. people spend a lot of time on the social media sites and read the various type of content there. This raises a question that is it influential in any aspect.
Thesis Statement: The social media platforms impact human behavior and their opinion about different things.
Section Heading: Research Question
Question # 1: Do social media affect the opinion of people regarding different aspects?
Question # 2: Do social media affect the opinion of the people positively?
Question # 3: Do social media affect the opinion of the people negatively?
Question # 4: What are the factors that lead to social media influence?
Literature Review
Section Heading: Social Media and Opinion of People
This section of this paper discusses the detail description and definition of social media and people opinion.
Social media is a platform that allows people to connect with each other and share their opinions and thoughts. Social media is in the form of websites and applications. In social media sites, there is content regarding different topics and problems are posted by different people. This content includes social issues, entertainment, history, trend, fashion, events, personal life, and different problems (Lee, 2016).
Section Heading: Connection between Social Media and Opinion of People
This section of the paper focuses on the connection among the social media sites and applications and the opinion of the people that how they are related to each other and what factors of social media impact people opinions.
There are 3.48 billion people in the world who use social media platform to be connected ("Digital 2019: Global Internet Use Accelerates - We Are Social", 2019). According to a survey, people use to spend approximately 2 hours and 22 minutes a day on social media sites and application, and those people are of all ages. Hence, it is evident that there is some connection between people and the social media sits. Behavior, opinion, and attitudes of people are changes on the basis of their experiences, and social media is a place where people get to know and experience different things, issues, topics on different aspects such as politics, social life, history, entertainm.
Sentiment analysis of comments in social media IJECEIAES
Social media platforms are witnessing a significant growth in both size and purpose. One specific aspect of social media platforms is sentiment analysis, by which insights into the emotions and feelings of a person can be inferred from their posted text. Research related to sentiment analysis is acquiring substantial interest as it is a promising filed that can improve user experience and provide countless personalized services. Twitter is one of the most popular social media platforms, it has users from different regions with a variety of cultures and languages. It can thus provide valuable information for a diverse and large amount of data to be used to improve decision making. In this paper, the sentiment orientation of the textual features and emoji-based components is studied targeting “Tweets” and comments posted in Arabic on Twitter, during the 2018 world cup event. This study also measures the significance of analyzing texts including or excluding emojis. The data is obtained from thousands of extracted tweets, to find the results of sentiment analysis for texts and emojis separately. Results show that emojis support the sentiment orientation of the texts and those texts or emojis cannot separately provide reliable information as they complement each other to give the intended meaning.
THE SURVEY OF SENTIMENT AND OPINION MINING FOR BEHAVIOR ANALYSIS OF SOCIAL MEDIAIJCSES Journal
Nowadays, internet has changed the world into a global village. Social Media has reduced the gaps among
the individuals. Previously communication was a time consuming and expensive task between the people.
Social Media has earned fame because it is a cheaper and faster communication provider. Besides, social
media has allowed us to reduce the gaps of physical distance, it also generates and preserves huge amount
of data. The data are very valuable and it presents association degree between people and their opinions.The comprehensive analysis of the methods which are used on user behavior prediction is presented in this paper. This comparison will provide a detailed information, pros and cons in the domain of sentiment and
opinion mining.
A sentiment analysis model of agritech startup on Facebook comments using na...IJECEIAES
Facebook page is a tool able to generate perceptions and acceptance, and support people and investors in making business decisions. Moreover, Facebook page plays a part in engaging people in the form of a community. People share experiences and opinions toward products, services, and trends in particular periods on the Facebook page community. Regarding sentiment analysis on Facebook pages, most education and other general topics in English have only been analyzed in English. However, sentiment analysis regarding agritech startups topics in Thai language has not been done yet. This study analyzes opinions and categorizes positive and negative comments by using naive Bayes classifier to examine the sentiments and attitudes of people and investors. The results could possibly reflect the perception rate of agritech startups in Thailand and could be applied to explain attentiveness and assess people’s engagement opinions. Furthermore, it could be applied in studying consumer behavior, marketing analysis, spread of information, and attitudes. The study's model is generic and could be applied in other contexts to provide insightful suggestions.
Social Media Influence Analysis using Data Science TechniquesMuhammad Bilal
The major purpose of this literature search report is to demonstrate the usage of different tactics of data science to investigate impact of social media while considering the interaction between influences and their followers.
IEEE PROJECTS 2016 - 2017
1 crore projects is a leading Guide for ieee Projects and real time projects Works Provider.
It has been provided Lot of Guidance for Thousands of Students & made them more beneficial in all Technology Training.
Project Domain list 2016
1. IEEE based on datamining and knowledge engineering,
2. IEEE based on mobile computing,
3. IEEE based on networking,
4. IEEE based on Image processing,
5. IEEE based on Multimedia,
6. IEEE based on Network security,
7. IEEE based on parallel and distributed systems
Project Domain list 2016
1. IEEE based on datamining and knowledge engineering
2. IEEE based on mobile computing
3. IEEE based on networking
4. IEEE based on Image processing
5. IEEE based on Multimedia
6. IEEE based on Network security
7. IEEE based on parallel and distributed systems
ECE IEEE Projects 2016
1. Matlab project
2. Ns2 project
3. Embedded project
4. Robotics project
5. IOT Projects
Eligibility
Final Year students of
1. BSc (C.S)
2. BCA/B.E(C.S)
3. B.Tech IT
4. BE (C.S)
5. MSc (C.S)
6. MSc (IT)
7. MCA
8. MS (IT)
9. ME(ALL)
10. BE(ECE)(EEE)(E&I)
TECHNOLOGY USED AND FOR TRAINING IN
1. DOT NET
2. C sharp
3. ASP
4. VB
5. SQL SERVER
6. JAVA
7. J2EE
8. STRINGS
9. ORACLE
10. VB dotNET
11. EMBEDDED
12. MAT LAB
13. LAB VIEW
14. Multi Sim
CONTACT US:-
1 CRORE PROJECTS
Door No: 66 ,Ground Floor,
No. 172, Raahat Plaza, (Shopping Mall) ,Arcot Road, Vadapalani, Chennai,
Tamin Nadu, INDIA - 600 026
Email id: 1croreprojects@gmail.com
website:1croreprojects.com
Phone : +91 97518 00789 / +91 7708150152
Industrial Training at Shahjalal Fertilizer Company Limited (SFCL)MdTanvirMahtab2
This presentation is about the working procedure of Shahjalal Fertilizer Company Limited (SFCL). A Govt. owned Company of Bangladesh Chemical Industries Corporation under Ministry of Industries.
Cosmetic shop management system project report.pdfKamal Acharya
Buying new cosmetic products is difficult. It can even be scary for those who have sensitive skin and are prone to skin trouble. The information needed to alleviate this problem is on the back of each product, but it's thought to interpret those ingredient lists unless you have a background in chemistry.
Instead of buying and hoping for the best, we can use data science to help us predict which products may be good fits for us. It includes various function programs to do the above mentioned tasks.
Data file handling has been effectively used in the program.
The automated cosmetic shop management system should deal with the automation of general workflow and administration process of the shop. The main processes of the system focus on customer's request where the system is able to search the most appropriate products and deliver it to the customers. It should help the employees to quickly identify the list of cosmetic product that have reached the minimum quantity and also keep a track of expired date for each cosmetic product. It should help the employees to find the rack number in which the product is placed.It is also Faster and more efficient way.
Immunizing Image Classifiers Against Localized Adversary Attacksgerogepatton
This paper addresses the vulnerability of deep learning models, particularly convolutional neural networks
(CNN)s, to adversarial attacks and presents a proactive training technique designed to counter them. We
introduce a novel volumization algorithm, which transforms 2D images into 3D volumetric representations.
When combined with 3D convolution and deep curriculum learning optimization (CLO), itsignificantly improves
the immunity of models against localized universal attacks by up to 40%. We evaluate our proposed approach
using contemporary CNN architectures and the modified Canadian Institute for Advanced Research (CIFAR-10
and CIFAR-100) and ImageNet Large Scale Visual Recognition Challenge (ILSVRC12) datasets, showcasing
accuracy improvements over previous techniques. The results indicate that the combination of the volumetric
input and curriculum learning holds significant promise for mitigating adversarial attacks without necessitating
adversary training.
Water scarcity is the lack of fresh water resources to meet the standard water demand. There are two type of water scarcity. One is physical. The other is economic water scarcity.
block diagram and signal flow graph representation
POLITICAL OPINION ANALYSIS IN SOCIAL NETWORKS: CASE OF TWITTER AND FACEBOOK
1. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
DOI : 10.5121/ijwest.2021.12401 1
POLITICAL OPINION ANALYSIS IN SOCIAL
NETWORKS: CASE OF TWITTER AND FACEBOOK
Telesphore TIENDREBEOGO and Yassia ZAGRE
University of Nazi Boni, Bobo Dioulasso, Burkina Faso
ABSTRACT
The 21st
century has been characterized by an increased attention to social networks. Nowadays, going 24
hours without getting in touch with them in some way has become difficult. Facebook and Twitter, these
social platforms are now part of everyday life. Thus, these social networks have become important sources
to be aware of frequently discussed topics or public opinions on a current issue. A lot of people write
messages about current events, give their opinion on any topic and discuss social issues more and more.
The emergence and enormous popularity of these social networks have led to the emergence of several
types of analysis to take advantage of them. One of them is the analysis of opinions in texts. It aims at
automatically classifying opinions in order to position them on a sentiment scale, thus allowing to
characterize a set of opinions without having to rely on a human to read them. Currently, opinion analysis
offers us a lot of information related to public opinion, either in the commercial world or in the political
world. Many studies have shown that machine learning techniques, such as the support vector machine
(SVM) and the naive Bayes classifier (NB), perform well in this type of classification.
In our study, we first propose an approach for tracking and analyzing political opinions in social networks.
Then, we propose a trained and evaluated machine learning model for political opinion classification. And
finally, the study aims at setting up a web interface to collect and analyze in real time political opinions
from social networks
KEYWORDS
Opinion analysis, opinion mining, Naive Bayes, Support Vector Machine, text classification, social
networks, machine learning, polarity
1. INTRODUCTION
Lately, social networks have become deeply rooted in the lives of their users. According to
Hootsuite[1] in January 2020 there were 3.8 billion active users in the world of social networks.
On the same date, Burkina Faso had 1.6 million active users of social media with a growth of
37% from 2019 to 2020. As a result, users are not only expressing their opinions on any topic, but
also increasingly discussing social issues and political events These exchanges create a large
volume of potentially exploitable information.
The development of social networks, which can be considered as a democratization of
computing, generates a rich mine of exploitable data called “Big Data”. This data is critical to
decision making for many individuals and organizations[2]. For this purpose, new analysis
technologies, alternatives to polls, have indeed appeared in recent years within the media and
political fields. As a result, researchers have since tried to conceptualize methods to analyze this
“Big Data” in a methodical, structured and efficient way[3]. Thus, among many methods, opinion
analysis has been developed to specifically analyze the polarity (i.e.: positive or negative) of texts
2. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
2
posted online on various levels. The objective of this type of analysis is generally to classify texts
or parts of texts according to the polarity of their author with respect to a given topic.
Our objective in this paper is to propose on the one hand an approach to monitor and analyze
political data generated in social networks. On the other hand, we propose a machine learning
model to analyze political opinions from social networks.
Our paper is structured as follows: we first present related work on opinion analysis. We then
provide an overview of techniques and applications. Finally, we present and explain the
implementation of a machine learning model for the analysis of our data. We also offer a web
interface to collect and analyze political opinions from social networks in real time.
2. RELATED WORKS
2.1. Social Networks
Social networks such as Twitter and Facebook have been the subject of a lot of research. Several
main factors explain this interest. The first is that these networks are almost entirely composed of
user-generated content. This would mean that every message posted on these sites has an opinion
or subjectivity of some kind. Moreover, the user is free to post whatever he or she wants, which
generates very diverse themes and topics, most of which are related to current events. In [2], the
authors point out that this portrait of opinion at a specific moment in time can be used to analyze
the evolution of opinion on certain subjects, provided that a periodic analysis of the latter is
performed. Secondly, the majority of social networks make their APIs available which facilitates
data collection as mentioned by Fang and Zhan (2015) [4]. Finally, the number of active people
online, who generate content, is massive [5]. Such social networks are therefore real data mines
that allow researchers, individuals as well as large companies to collect data among a wide range
of topics.
2.2. Social Networks and Political Opinions
The potentials of social media seem to be the most promising in the political context as they can
be a catalyst for more participation and democracy [6]. In [7], Stefan Stieglitz defines public
participation as the process by which the concerns, needs, and values of the public are
incorporated into the decision-making of governments and corporations. In a similar vein,
Karpf[8] introduces the notion of “Political 2.0” which can be understood as exploiting the
reduced transaction costs of the Internet and its abundance of information, in order to create more
participatory and interactive political institutions. There is a large body of research regarding the
role of social networks in political thinking. The US presidential campaign revealed that social
networks have become increasingly essential to political communication and persuasion. Recent
studies have shown an emerging need for political institutions as well as government departments
to leverage the resources of social networks to improve services and communication with citizens
and voters [8]. Similarly, Williams and Gulati [9] assert that it has become increasingly important
to stay abreast of ongoing discussions and manage one’s own reputation in virtual communities,
especially with respect to emerging topics that may lead to scandal or crisis for a specific
politician or party. In a survey of members of the German “Bundestag”, Stieglitz and Stefan [8]
found that most members of parliament would like to benefit from concepts and instruments to
identify sensitive political topics early on. However, this task requires a lot of effort, appropriate
tools and, above all, a systematic approach.
3. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
3
2.3. Opinion Analysis
Interest in the subject of sentiment analysis seems to be increasing year after year since 2002. At
least that’s what we could tell by looking at the evolution of the search for the phrase “sentiment
analysis” in the google search engine. See Figure 1.
Figure 1: Evolution of searches made with the expression sentiment analysis in the Google search engine
since 2002in the world
It seems that the attraction for this topic is also shared by the research world. Indeed, more than
7000 articles have been published on this subject until 2019[10]. The analysis of the opinion thus
occupies a place of a certain importance from the interest that is granted to it and by all the
literary activity that is dedicated to it. In the literature, sentiment analysis, opinion mining,
opinion extraction, sentiment mining, subjectivity analysis, emotion analysis, review mining, are
terms used to designate technologies for the automatic analysis of written or spoken discourse in
order to extract subjective information such as judgments, evaluations or emotions.
According to Cambria[11], opinion analysis has been one of the most active research areas in
natural language processing since the early 2000. The purpose of opinion analysis is to design
automatic tools to extract information from a natural language text, such as opinions and feelings,
in order to generate knowledge that can be used by a decision support system or by a decision
maker. Opinion analysis is the computer processing of a text that is explored in order to identify
the general sentiment or opinion expressed. We often speak of classifying the polarity (positive,
negative or neutral) of a text. The exploration is done on an entity and these entity can represent
individuals, products, events or various subjects that are likely to be covered by comments,
criticisms or appreciations.
In order to identify the polarity of a text, opinion analysis relies on several features which are
among others: the frequency of a term, the importance of a term in the text, negations in
sentences [12]. In other words, opinion analysis concerns the processing of an opinion text to
extract and categorize the opinions of a certain document. The polarity of the sentiments usually
expressed in terms of positive or negative opinion (binary classification).However, there can be a
multiple classification, where the feeling can have a neutral label or even a different label like
very positive, positive, neutral, negative, very negative, the labels can also be associated with
emotions such as sadness, anger, happiness, etc.
4. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
4
2.3.1. Opinion Analysis Levels
The research on opinion analysis was mainly conducted at three different levels of analysis [2]:
Message level or Document level, Sentence level and aspect level.
2.3.2. Opinion Analysis and Applications
According to Pang and Lee [2], it was in 2001 that the field of automatic opinion analysis began
to attract attention. The numerous possibilities brought by this method and the doors it could
open eventually aroused the interest of many political, sociological, governmental, financial,
advertising and commercial fields.
• Politics: Nowadays, political actors follow the trend of opinion analysis, because before
declaring a new law, politicians try to collect the opinion of social network users about this
law. Through opinion analysis, policy makers can collect citizens’ opinions on certain
policies, in order to benefit from this information to improve or create a new policy that
suits with citizens. It is highly strategic to also know the opinion of Internet users on a
politician during a presidential election [13][14]. It should be noted that people’s opinion
and experience is a very useful element in the decision-making process.
• Economy: Thanks to the analysis of opinion, the companies can know the opinion of the
customer son their products or their services. In a perspective of improvement of their
products and increase of their sales and incomes. Marketing has quickly understood the
interest of opinion analysis. An example of commercial interest found in Pang and Lee’s
study [2] is that of a marketer wanting to know why his new product was not selling well.
He was therefore interested in the opinions of his consumers regarding his product. The
result was useful for him because by knowing their opinions, he was able to orient his offer
according to the requirements of the market.
• Other areas of application: Recently, opinion analysis is also being applied in totally
different fields, namely on-profit fields. The two most sought-after domain sat the moment
are the detection of cyber stalking as well as the detection of suicidal tendencies of people
active on the web[15][16][17][18].
2.3.3. Opinion Analysis Approach
In the literature review, there are generally three approaches to opinion analysis:
Lexicon based approach;
Machine Learning based approach;
Hybrid approach
Lexical approach: According to Annett and Kondrak, Sasikala and Mary, the lexical approach
consists of identifying words present in a dictionary of pre-labeled words (each assigned a value
on a scale composed of several sentiment intensities) and tallying the sentiment numbers to arrive
at a total polar score for each text [19][20]. This approach therefore requires the availability of a
dictionary that is adapted to the language used in the texts understudy. Moreover, according to
Muhammad et al, for the dictionary to be effective, it must be adapted to the domain under study
[21].
Machine learning approach: A major shortcoming of this lexicon-based approach is its inability
to identify opinion words with domain- and context-specific orientations. As, a sub-field of
artificial intelligence (AI), machine learning based approach refers to the techniques and actions
5. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
5
that researchers have to develop an automatic system that, by analyzing data, learns by itself.
These learning algorithms allow systems to train themselves with the data provided in order to
develop and improve. From a set (or model) of data called “dataset”, the algorithms are trained
and tested. the algorithms are trained and can, from these models, make predictions about other
data they may encounter in the future. The machine learning approach is subdivided into
supervised, unsupervised and semi supervised learning methods.
Hybrid Approach: This approach combines the lexicon based approach and the machine learning-
based approach and tries to correct the drawback of the lexicon-based approach of low recall and
the machine learning-based approach of manual annotation.
Classification algorithms: There are different classifiers in machine learning methods that are
used for opinion analysis. The Support Vector Machine method was introduced by Joachims
[22][23], then used by Drucker [24], Taira and Haruno [25], and Yang and Liu[26]. The Naïve
Bayes classifier is a supervised probabilistic learning technique that is regularly used for
classification purposes. It is based on Bayes’ theorem where the conditional probability of an
event X occurring, for an index Y, is determined by Bayes’ rule [2] (See equation 1).
P (X/Y ) = P(X)P(Y/X)/P(Y ) (1)
The k-Nearest Neighbors KNN method is a classification method in which the model stores
observations from the training set for classification of the test set data. A text classifier based on
the decision tree method is a tree of internal nodes that are labeled with terms, the branches
coming out of the nodes are tests on the terms, and the leaves are labeled with categories
[27].Neural networks constitute a network of units, where input units correspond to terms, output
units to categories of interest, and weights of edges connecting units to dependency relationships.
To rank a test document d_i, its weights W_ki are loaded into the input units; activation of these
units propagates through the network, and the value of the output unit(s) determines the ranking
decision [4].
2.3.4. Opinion Analysis Problems
In the literature, opinion analysis is a very complex task with many unsolved problems[2]. The
following are some of the challenges in dealing with this area. The handling of comparisons and
the handling of negation, the detection of false opinions[2], the problem of language, Aboutness,
the detection of sarcasm[28].In this section, we have first presented what opinion analysis is, with
its three different levels of analysis and its applications. Then, we presented the approaches that
exist in the literature with the different algorithms.
And finally, we evoked according to the literature the problems which existed in this type of task.
Studies have shown that social networks such as Facebook and Twitter are a reliable indicator of
the general opinion of users on entities, events or institutions, etc.
3. METHODOLOGY
In this section, we propose an approach to tracking and analyzing political opinions from Twitter
and Facebook. While data tracking and monitoring concerns different approaches to how and
what kind of politically real data generated by users of different social network platforms can be
tracked and collected, the section on data analysis discusses different methods of analysis for
different purposes and approaches.
6. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
6
3.1. Data Tracking Approach
Tracking sources: The first step in data collection is to determine the sources of the data. For this
reason, we decided to develop our approach by focusing on Twitter and Facebook as the primary
sources of social media detain the case of Twitter, the data to be monitored and exploited is in the
form of public "tweets" that are easily accessible. For Facebook, the most important type of data
is the content of the “wall”, including “status updates” and related comments. It is important to
note that data on Facebook is, unlike Twitter, semi-public in nature,i.e., data can only be
collected from public profiles (or more concretely, groups or “pages”, since most individual
profiles are not public). Specifically, in a political context, a list of politically relevant Facebook
groups and pages should be predefined as data sources. Tracking methods: Both Twitter and
Facebook offer application programming interfaces (APIs) for tracking data. The most frequently
used APIs for Twitter are the Search API and the Streaming API, while Facebook’s Graph API
allows programmers to conveniently track wall posts (i.e. status updates and comments).
Tracking Approaches: We have identified four approaches to tracking political data, which
depend on the specific intentions of the tracker: (1) self-reputation,(2) keyword/topic-based, (3)
actor-based, and (4) random/exploratory.
self-reputation Approach: The first approach is applicable when, for example, politicians
or political parties want to know explicitly how people talk about them in social networks.
In this case, politicians or parties can have all tweets that contain their name collected,
either as a simple keyword or as a hashtag. If they have their own presence on Facebook in
terms of a page or group, they should follow all posts and corresponding comments
published by users or members of their own page or group.
Keyword/Topic-based Approach: Politicians are often very interested in the reactions or
opinions of Internet users on certain political topics. In this case, the second tracking
method seems appropriate, as it tracks Tweets as well as Facebook posts that contain
keywords related to the topics of interest. To achieve a high level of data completeness, the
relevant keywords of the topic of interest must be carefully selected. The more important
the topic to be analyzed, the higher the number of keywords to be taken into account.
Actor-based Approach: In political communication, a number of actors can generally be
considered more influential or popular than most other users. These actors are said to have
the power to influence opinion-forming processes (online)[2]. Therefore, politicians or
political parties are also interested in tracking these important users in terms of generated
content. For this, an actor-based tracking approach could be used to track tweets, wall
posts as well as corresponding comments specifically provided by these influential users
who should also be identified in advance.
Random/Exploratory Approach: Unlike the three tracking approaches that are more
targeted in nature, the fourth approach supports content mining. The principle of this
tracking approach is to randomly determine one or more data sets (tweets or Facebook
posts) for different time periods for analysis. Based on this random data, a content analysis
can be used to identify key political topics and detect the opinions or feelings of users
associated with these topics.
3.2. Opinion Analysis Approach
A major drawback of the lexicon-based approach is its inability to identify opinion words with
domain- and context-specific orientations. Based on the difficulties of the lexicon-based method
reported in the literature review, it is appropriate to use a machine learning technique. For this
purpose, we have chosen machine learning algorithms according to their popularity in the
7. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
7
literature and according to their predictive potential in the context of opinion analysis. We chose
four algorithms, including support vector machines (SVM), decision tree (DT), random forest
(RF) and Naïve Bayes (NB). Random Forest (RF) and Naïve Bayes (NB). Finally, we compared
the classification results of the four algorithms in order to select the one that gives the best result
in the classification of political opinions.
3.3. Data Representation
The data representation methodology that we adopt takes into account the information that was
synthesized in our literature review. Indeed, we have chosen to prepare the data according to the
types of representations and transformations that we thought were the best given the
characteristics of our data, namely Term Frequency Inverse Document Frequency (TF-IDF).
Considering the terms j in a document i being part of a certain corpus of documents, here is how
we could represent the TF-IDF under an algebraic form[4] (see equation 2).
𝑽𝒊𝒋 = 𝑻𝑭𝒊𝒋 . 𝑰𝑫𝑭𝒋 such as 𝑻𝑭𝒊𝒋 =
𝒇𝒊𝒋
𝒏𝒊
𝑎𝑛𝑑 𝑰𝑫𝑭𝒋 = 𝒍𝒐𝒈
𝒎
𝒃𝒊𝒋
𝒎
𝒊=𝟏
(2)
where:
𝑽𝒊𝒋 = the value of the term j in the document i.
𝒇𝒊𝒋 = number of times the term j in the document i.
𝒏𝒊 = total number of terms in the document i.
𝒃𝒊𝒋 =1 if the term j is present in the document i and 0 else.
𝒎= if the term j is present in the corpus.
3.4. Data Pre-Processing
The characteristics of tweets generally boil down to their limited length and the use of informal
language. Thus, Twitter users use abbreviations, emoticons, and slang to express their opinions
and feelings. Therefore, a pre-processing step is essential. In the following we present the
preprocessing procedure followed in our work, whose goal of this step is to clean up the tweets
and make them as close as possible to a formal language. First we started by eliminating URLs
and words with hashtag, non-alphabetic characters, repeated terms and punctuations.
3.5. System Architecture
Our study focuses on the task of opinion analysis on social network data in particular Twitter and
Facebook. To achieve this goal and to get the best possible performance, we propose the
following architecture. Figure 2 below illustrates the general process of the system which
includes several steps including data collection, pre-processing, annotation, sentiment analysis
and visualization.
8. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
8
Figure 2: System architecture
The collection phase scans Twitter and Facebook to extract tweets and comments thanks to the
API of the social network. The result of the collection goes through a step of pre-processing
which allows to remove the duplicates and the spams (same author, same content, same post).
Once the data has been pre-processed, comes the annotation phase, which allows to create a
training dataset for the opinion analysis system. The text stream from the comments is then sent
to categorization to determine the sentiment class, which in turn uses the data from the annotation
layer for learning. The visualization phase draws from the latter to expose the results to the user
according to the desired parameters. The final application will take the form of a Dashboard that
present she aggregated statistics of the analysis. In this section, we have proposed an approach to
tracking and analyzing political opinions from social media. Specifically, we have described
various data tracking approaches and corresponding analysis methods that could help to better
understand political discussions in socialmedia. From a practical perspective, this approach
should serve as a guideline for developing tools to collect, store, analyze, and summarize user-
generated social media content that is politically relevant to political institutions. From a research
perspective, the framework provides a comprehensive overview of different specific
methodological approaches that can be used to analyze political opinions from social networks.
4. BACKGROUND
These first years of the 21st century seem to be marked by multiple changes in the production and
exchange of information. Thus, since the advent of Web 2.0 and especially the massive use of
social networks, the number of users has considerably increased and their role has evolved into
that of information provider. As a result, an increasing number of users send hundreds of millions
of messages every day to specially created websites, such as Facebook and Twitter. Users write
about their vision of society, increasingly discuss social issues and political events, share
opinions and divulge ideas [4].
As a result, social networks have become an essential research material for researchers in
computer science and the humanities. To this end, new analysis technologies, alternatives to polls
and surveys which are reputed to be long and costly, have indeed appeared in recent years within
the media and political fields. Thus, among many methods, opinion analysis has been developed
to specifically analyze the polarity of texts. Hence the growing development of techniques to
capture these evaluations of Internet users, ranging from the simple counting of positive or
negative comments to the more detailed analysis of the contents of these comments. It is in this
9. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
9
context that this paper is written. The objective of this paper is to propose an approach to monitor
and analyze political data generated in social networks. On the other hand, we propose a machine
learning model to analyze political opinions from social networks.
5. RESULTS AND DISCUSSION
In this section, we first present the data that was used to train our model. The input data we used
are tweets, extracted from a dataset to train and test our model. We also test the proposed model
in real time on tweets extracted from Twitter itself. These tweets represent the statuses posted on
this social network and related to the political field.
5.1. Dataset
To train our algorithms, it is important to have well annotated data. To do so, we thought it wiser
to take texts already marked by their authors rather than marking +9them ourselves. For the sake
of rigor, we also considered the size of the sample as an essential factor to improve the value of
the results. Our choice was therefore theSentiment140 dataset which is part of the
Sentiment140[29] project at Stanford University. Indeed, this database contains thousands of
tweets with their respective ratings, and therefore has the necessary assets for the application of
our classification. It is a large dataset, more than one million tweets with two classes of negative
and positive tweets, with 800,000 for each. The authors represented the positive class with the
number 4 and the other class with the number 0.
5.2. Presentation of Results
Recall that the objective of our study is to implement a system to classify tweets or comments on
political topics. Such a system consists of two phases in the implementation. The first step is a
machine learning model trained with algorithms and among them the one that performs best. The
second phase is the implementation of a web application using a machine learning model and
allowing to visualize the trends in real time. We will present here the results of the machine
learning models.
For the validation of the performances of our model, we use the 80% 20% method, as 80% used
in the learning phase, and 20% for the testing phase. In agreement with [30]. Table I presents the
confusion matrix that applies to our binary classification context. The performance measures used
are precision, recall and F1-measure, whose bases of calculation are relative to the confusion
matrix.
Table 1 : Confusion matrix in a polar classification contexte.
Positive Negative
Positive True Positive (TP) False Positive (FP)
Negative False Negative (FN) True Negative (TN)
The measurements are presented in the following formulas:
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 =
𝑇𝑃
𝑇𝑃+𝐹𝑃
(3)
This measure represents the proportion of correct predictions to the total number of predictions.
predictions.
10. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
10
Recall =
TP
(TP + F N)
(4)
This metric corresponds, for a binary classification, to the number of positive predictions of the
model out of the total of predictions that should be positive. Otherwise, it is the proportion of
well-ranked items to the number of items in the class being predicted.
To combine all this information, we often introduce a metric called the F1-Score. It is calculated
in this way.
𝐅𝟏 − 𝐬𝐜𝐨𝐫𝐞 =
1
1
1
1
𝑅𝑒𝑐𝑎𝑙𝑙
+
1
𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦
= 2 ∗ 𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 ∗ 𝑅𝑒𝑐𝑎𝑙𝑙(𝐴𝑐𝑐𝑢𝑟𝑎𝑐𝑦 + 𝑅𝑒𝑐𝑎𝑙𝑙(4)
Represents the trade-off measure between precision and recall. In the table II, we summarize the
result obtained for each model.
Table 2 : The validation results of dataset.
Algorithms NB SVM RF DT
Accuracy 83.26% 82.75% 69.78% 77.08%
Recall 0.7 0.8 0.6 0.7
F1-score 0.7 0.8 0.6 0.7
CPU time 1mn 50s 4h 54mn 30s 8mn54s 2h 56mn 55s
Our dataset consisted of nearly 1.6 million tweets evenly distributed in the 2 categories. In order
to select the best algorithm, we used several different algorithms. Thus, after running several tests
with different models and configurations, we selected Naïve Bayes as the best performing
classifier for this particular problem. To evaluate the results, we used the 5-way cross-validation
method and our best performing classifier reaches an accuracy of 83.26%.
6. CONCLUSIONS
The arrival of the digital age, coupled with the rapid expansion of the web and its
democratization, has profoundly changed the way society works. Now, the Internet is full of user-
generated content. This content, called “Big Data”, has become an invaluable mine of data that
can be exploited in many financial, political and no lucrative fields. At the same time, a number
of methods to successfully analyze this data have emerged. Among these methods, opinion
analysis has been developed to specifically analyze the polarity of texts posted online on various
levels. These methods have become even more relevant due to the rise in popularity of social
networks. The leaders in this field, Facebook and Twitter, have billions of users who are
constantly generating new content while expressing their opinions.
In this paper, we focused on the analysis of political opinion in social networks. To achieve our
goal, we proposed an approach for tracking and analyzing political opinion from social networks.
Specifically, we have described various approaches to data tracking and the corresponding
analysis methods On the other hand, we trained a machine learning model for political opinion
analysis from social networks. To do so, we used four machine learning algorithms (SVM, DT,
11. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
11
RF, NB) that we trained with the sentment140 [36] dataset. At the end of the training, we chose
NB which is the best performing classifier in terms of accuracy. It reaches an accuracy of83.26%.
Ultimately, the results obtained demonstrate that social networks, and Twitter and Facebook in
particular, are a source of data for obtaining key indicators for better decision-making.
However, we could agree with some authors[3][4][12]that the opinions coming from social
networks would not constitute a fairly representative sample of the distribution of the whole
population, especially in our countries where the digital penetration rate remains very low, i.e.
22%penetration rate in Burkina Faso with 1. 6 million active users of social networks in January
2020 [31]. This could also be a significant bias. Nevertheless, we are witnessing a very rapid
evolution of the penetration rate. This is the case in Burkina Faso, which has seen a strong 35%
grow thin active social media users from one year to the next [31].
As perspectives to our work, we foresee:
test our model on other data sets;
develop the model to manage the problem of sarcasm;
Finally, we believe that it is necessary to define a framework that will allow an efficient analysis
of the data from social networks in order to derive several useful information for decision
making.
REFERENCES
[1] Hootsuite, «Digital marketing and management,» janvier 2020. [En ligne]. Available:
https://www.hootsuite.com/.
[2] L. L. a. B. Pang, «Opinion Mining and Sentiment,» in Foundation and Trends in information
Retrieval, p. 63, 2008.
[3] D. B. a. A. Lohard, «Opinion mining and Sentiment analysis,» HEC MONTREAL, 2015.
[4] J. Labrecque, «Analyse des sentiments dans les textes à l’aide d’algorithmes d’apprentissage
automatique,» HEC MONTREAL, 2015.
[5] F. X. a. Z. Justin, «Sentiment analysis using product review data,» Journal Of Big Data, 2015.
[6] A. M. K. a. M. Haenlein, «Users of the world nite ! The challenges and opportunities of social
media,» Business Horizons, 2009.
[7] A. M. K. a. M. Haenlein, «Users of the world unite ! The challenges and opportunities of social
media,» Business Horizons, 2009.
[8] D. A. Karpf, «Recherche sur la blogosphère : une approche à méthodes mixtes pour des systèmes en
évolution rapide,» IEEE Intell Syst 24, 2009.
[9] C. B. a. G. J. Gulati, «Facebook grows up: An empirical assessment of its role in the 2008
congressional elections,» In Annual Meeting of the Midwest Political Science Association, 2008.
[10] F. Ronen, «Techniques and applications for sentiment analysis,» Commun.ACM, 2018.
[11] C. Erik, «Affective computing and sentiment analysis,» IEEE Intell, 2016.
[12] L. Bing, «Opinions, Sentiment, and Emotion in Text,» Cambridge University University Press, 2015.
[13] B. F. a. Sabatier, «Expérimentation de fonctions pour la détection d’opinions dans les Blogs,» Institut
de Recherche en Informatique de Toulouse, 2010.
[14] E. C. a. D. O. a. D. Rajagopal, «SenticNet 3:a common and common-sense knowledge base for
cognition driven sentiment analysis,» In Twenty-eighth AAAI conference on, 2014.
[15] V. H. a. C. L. a. E. B. V. a. J. M. a. B. D. a. D. G. P. a. V. Hoste, «Detection and finegrained
classification of cyberbullying events,» In International In International, 2015.
[16] B. D. a. V. Hoste, «Online suicide prevention through optimised text classification,» Information
Sciences, 439-440, 61-78., 2018.
[17] J. M. X. a. X. Z. a. A. Bellmore, «Fast learning for sentiment analysis on bullying,» In Proceedings of
the First International Workshop on Issues of Sentiment Discovery and Opinion, 2012.
12. International Journal of Web & Semantic Technology (IJWesT) Vol.12, No.4, October 2021
12
[18] J. M. Gómez, «Language technologies for suicide prevention n social media,» In Proceedings of the
First International Workshop on Issues of Sentiment Discovery and Opinion, 2014.
[19] M. A. a. G. Kondrak, «A comparison of sentiment analysis techniques : polarizing movie blogs,» In
Proceeding of the Canadian Society for computational studies of intelligence, 21st conference on
Advances in artificial intelligence., 2008.
[20] S. P. a. L. Mary, «Comparative Study of Sentiment Analysis Techniques in Web,» Immaculate
Sheela. International Journal of Scientific Engineering Research, 2017.
[21] A. M. a. N. W. a. R. L. a. R. Glassey, «Domain-based lexicon enhancement for sentiment analysis,»
2017.
[22] J. Torsten, «Text categorization with support vector machines : learning with many relevant
features,» 10th Euro pean Conference on Machine Learning, ECML-98, 1998.
[23] J. Torsten, «Transductive inference for text classification using support vector machines,,» 16th
International Conference on Machine Learning, ICML-99, 1999.
[24] H. D. e. V. V. e. D. Wu, «Automatic text categorization and its applications to text retrieval,» IEEE
Trans. Neural Netw, 1999.
[25] H. T. e. M. Haruno, «Feature selection in SVM text categorization,» 16th Conference of the
American Association for Artificial Intelligence, AAAI-99, 1999.
[26] Y. Y. a. X. Liu, «A re-examination of text categorization methods,» 22nd ACM International
Conference on Research and Development in Information Retrieval, 1999.
[27] T. Mitchel, «Machine Learning,» McGraw Hill, 1996.
[28] M. Abdelkader, «L’analyse du sentiment utilisant le deep Learning,» Université Dr. TAHAR
MOULAY SAIDA, 2019.
[29] S. university, «trainingandtestdata,» 2012.
[30] M. S. e. N. J. e. S. Szpakowic, «Beyond accuracy, f-score and roc : A family of discriminant
measures for performance evaluation,» In Australian Conference on Artificial Intelligence,volume
4304 de Lecture Notes in Computer Science, 2006.
[31] lekiosquedigitalduburkina.com, «lekiosquedigitalduburkina,» Business Horizons, 2018. [En ligne].
[Accès le janvier 2020].
[32] V. Antoine, «L’analyse des sentiments dédiée aux weets politiques : une approche par apprentissage
automatique supervisé,» Aantal woorden:19.168, p. 19, 2017-2018.
AUTHOR
Yassia ZAGRE, Master degree in information systems and decision support system
in nazi BONI university Burkina Faso. My interest research topic is image
watermarking for decision support and multimedia system.
Telesphore TIENDREBEOGO, PhD and overlay network and assistant
professor at Nazi Boni University. I have a master's degree in multimedia and real
time system. My current research is on big data and image watermarking