Belonging Bot

B
e
l
o
n
g
i
n
g
 

B
o
t

A means to understand social connectedness online
through folksonomies & sentiment analysis on Twitter.

T
w
i
t
t
e
r
 

S
t
a
t
i
s
t
i
c
s

21 Followers

101 Following

192 Tweets

265 Likes

W
h
o
 

i
s
 

B
e
l
o
n
g
i
n
g
 

B
o
t
?
 

Belonging Bot is a Twitter Bot that interacts with other Twitter users that are discussing topics such as: Loneliness, Belonging and the Lockdown in South Africa. It does this to try to understand social connectedness online further, by analysing Twitter interactions with the help of Folksonomies and Sentiment Analysis!

Once it understands the experiences of different communities in South Africa during this time, my bot will try to understand how and why conversations between users areperpetuated. Belonging Bot attempts to encourage the sharing of positive posts by responding to and retweeting them - adding imagery and emojis to include a visual break from all of the text that Twitter is usually made up of.

B
e
l
o
n
g
i
n
g
 

B
o
t
'
s
 

F
i
n
d
i
n
g
s

Sentiment Analysis

Which sentiments are most common for the chosen folksonomies?


All

Because Belonging Bot is aiming to create a positive space on Twitter, the majority of the tweets that it has interacted with and collected are positive. Neutral tweets have been collected, as it is often news sources sharing these, and negative tweets have been interacted with, as Belonging Bot aims to further understand social connectedness and, in doing so, has found that the users posing negative tweets are often aiming to find a sense of social connectedness online.

Types of Interactions

How are users interacting within different folksonomies on Twitter?

Belonging Bot's Interactions

The act of exploring another user’s page, and finding similar interests and experiences could be a factor in gaining a feeling of social connectedness. Maslow’s Third Need states that people will aim to form part of a group. One of the first steps to being included in a group is to share a similar interest. By having users click on Belonging Bot’s page, it can be assumed that they found a post interesting and relevant to themselves and chose to explore more. Throughout this practical project, I have found that Belonging Bot received more impressions when its own, original content was posted, as opposed to when it reposted others’ tweets.

Belonging Bot's Engagements

Belonging Bot gained more followers when it began posting statuses of its own, as opposed to just sharing other users’ posts. This is evident with an increase of 829% profile visits in the 28 days since Belonging Bot began posting statuses of its own. These statuses consisted of extracts from my research question, quotes from sources that I researched, and emoticons such as hearts to attempt to brighten up Belonging Bot’s profile and allow it to seem welcoming. Another insight of social connectedness on Twitter that I was able to draw from Belonging Bot’s interactions is that reaching out to other users will often result in the user reciprocating the effort.

User's Interactions

How are users accessing the Folksonomies on Twitter, and what are they posting?

Device

According to Hootsuite “a social media management platform” (Simpson 2020), South Africa’s social media usage increased by 50% in the first two months after the lockdown was implemented, with Twitter’s usage increasing by 14.09%. When looking at device ownership of internet users aged between 16 and 64, 94% own mobile phones, compared to 76% who own laptops or desktop computers. The web traffic of mobile phones is 73.4% compared to only 24.5% for laptops and desktops. this shows that more users in South Africa are connecting to the internet, and by extension social media, on their mobile devices. Hootsuite reported that 96% of the users aged between 16 and 64 use social networking mobile applications. Twitter falls within this realm.

Content Type

The majority of posts within my chosen folksonomies are users’ own posts. this means that these are the original thoughts all tweets of the users interacting with them these folksonomies (#lockdownSouthAfrica, #loneliness, #belonging). This could prove that the tweets being sent out are more personal and involve users sharing their own personal details and experiences. The second-highest type of post is replies. this indicates that conversations are forming around these folksonomies. use this could be replying to one another's posts in order to find out more information, offer support, or simply express that they can relate. Lastly, users are sharing one another's posts. This could be in order to draw attention to specific instances that are taking place, or as a way to express that one relates to what the original user was saying in their post.

F
o
l
l
o
w
e
r
 

A
l
g
o
r
i
t
h
m

Belonging Bot provided me with insights with regards to how and why users visit a profile. The act of exploring another user’s page, and finding similar interests and experiences could be a factor in gaining a feeling of social connectedness. Maslow’s Third Need states that people will aim to form part of a group. One of the first steps to being included in a group is to share a similar interest. By having users click on Belonging Bot’s page, it can be assumed that they found a post interesting and relevant to themselves and chose to explore more.



if ( Follower Count < 30 && Sentiment Score < 0 ) {
     Users are more likely to connect with Belonging Bot
}

When analysing the tweets that Belonging Bot interacted with, accounts with more friends than followers often posted more tweets with a positive sentiment than negative. This could be as a result of the user feeling a sense of social connectedness, as they have found a community with which to interact on Twitter. The creations and interaction within this community satisfy Maslow’s third need, as interpersonal connections have been established. Interestingly, the majority of Belonging Bot’s followers were users who had posted tweets with negative sentiments and had Belonging Bot add them to a collection. This could show that users posting tweets with negative sentiments could be attempting to foster connections online and be aiming to form part of a community.



S
o
u
t
h
 

A
f
r
i
c
a
n
 

C
i
t
i
e
s

Main South African cities where different folksonomies are used



lockdownSouthAfrica


#lockdownSouthAfrica

#lockdownSouthAfrica was more widely shared in larger metropolitan areas.This could be for one of 2 reasons. The first is that these areas are more likely to have a large variety of news and media networks operating in them, meaning that information on COVID-19 and the societal restrictions that it has caused would be shared by a variety of different sources. These sources would use this folksonomy to ensure that the information is easily available and found. The second reason is the fact that these areas are more densly populated, so their citizens would be more used to interacting with one another and thus feel the lack of real-world social connectedness more than people who do not often interact with others.

D
e
m
o
g
r
a
p
h
i
c
s
 

Main South African cities where different folksonomies are used

An exploration of the genders, ages, cities and interests of the people who interacted within the folksonomies that I chose, and how they see themselves.

This will allow a better understanding into the types of people interacting in different Folksonomies, and why these Folksonomies may be relevant to them.



Find out more about the users who interactedin different Folksonomies →

P
o
p
u
l
a
r
 

H
a
s
h
t
a
g
s

Hashtags associated wiith different Folksonomies



lockdownSouthAfrica


#lockdownSouthAfrica



When looking at the hashtags that are associated with the folksonomy #lockdownSouthAfrica, we can see that #COVID19 is the most used, more popular one. This is because the lockdown was only implemented due to the pandemic. Another two hashtags that were very popular within this folksonomy are #staysafe and #stayhome. These were initiatives that were launched in order to get people to abide by the regulations set out in the lockdown. They show that users care about their fellow citizens, and are aiming to stop the spread of the disease.

D
e
f
i
n
i
t
i
o
n
s

Sentiment Analysis

Sentiment analysis is a way to treat “opinions, sentiments and subjectivity of text” through computational processes (Medha, Hassan & Korashy 2014:1). Thus, it is a way to discover people's opinions towards subject-matter online.

Folksonomies

A folksonomy is a way in which items on the internet are described and categorised by users (Neal 2007:7). This is done by attaching keywords to posts through the use of tags and hashtags, to structure and organise them.

Social Connectedness

In the article, “How Social Are Social Media? A Review of Online Social Behaviour and Connectedness”(2017) by Tracii Ryan, Kelly A. Allen, DeLeon L. Gray and Dennis M. McInerney, social connectedness is defined as feeling as though one belongs in the same group as, and relates to, one’s peers (2017:1).

W
h
a
t
'
s
 

N
e
x
t
?

With regards to future research, this study can be built upon to better incorporate social media users from different locations and within different folksonomies into online communities, by fostering a sense of social connectedness. Through this research, I discovered which users are reaching out in search of social connectedness in the form of conversations online. I aim to encourage online spaces to become more accepting of these users, and to better integrate them into online communities. Discovering specific folksonomies, as well as different geographic locations, in which users feel isolated or lonely could be done through surveys and interacting with more users. These users could come from a wider range of folksonomies, geographical locations, and social media platforms. By discovering the shortcomings within these spaces, social connectedness could be better fostered online and, by interacting within different folksonomies, and on different platforms, I would be able to see if the trends that I identified correspond.