🍞 bRead

🍞 What is the problem?

A Filter Bubble is a state of intellectual isolation that can result from personalized searches when a social media algorithm selectively guesses what information a user would like to see based on their interaction history and other data which platform knows about them.

It is difficult for a person to find opposing views on a topic on social media. One has to make a conscious effort to get diverse views on some topic, and break out from the filter bubble created by the social media algorithms. However, some choose to keep their filter bubble intact in order to prevent information overload which leads to misinformation, another problem prevalent in OSMs today.

🍞 What is bRead? 

bRead is a new outlook on interacting with social media where people view multiple sides of a Debate, Question or News Topic and contribute to the conversation, hence delivering discourse in an organized, ethical and user-centric function.

🍞 How did we create bRead?

Initial User Study

60 participants in the age range of 15-25 years were initially invited to fill a survey to test various hypotheses regarding how their usage of a social media platform is impacted by diversity of the social media feed, targeted advertisements, organization of the feed, etc.

Results from the survey provided considerable evidence for the following hypothesis: 

“Users prefer more organization and content diversity in their social media feeds than there is currently”.

We then conducted 2 sets of interviews with 10 participants: 6 Twitter users who have actively engaged on the platform for 4-5 years on average, and 4 Researchers whose work is based in the field of social media and filter bubbles. 

The users were asked a wide range of questions to gauge how they engage with content and other users on the platform, and their views on the platform’s design and the issue of Filter bubbles. Results from the User interviews are summarized as follows:

We then gauged the opinion of Researchers on what they perceive as Filter bubbles, and whether we should be done away with. The results from the interviews validate our hypothesis.

🍞 Enough Talking! Show me bRead!

The homepage provides an intelligently curated list of topics a.k.a. loafs with each loaf containing discourse around politics, entertainment, sports, global affairs or any relevant topic appealing to the masses.

But what does a loaf contain? Simple, Breads.

Each loaf contains a repository of posts a.k.a. breads segregated in two parallel views: breads in favor of the loaf and breads against the loaf. The users of the platform collectively decide the category of every bread.

But what empowers users to decide the category? We use Community Driven Annotations.

For every loaf, whenever any frequent visitor enters a loaf, they are considered as a knowledgeable person and asked to annotate if a bread is against the topic or in favour of the topic. 

But who is a knowledgeable person? We assign them a Credibility Score.

Every user starts with a baseline credibility score of 50%. If their annotation score is close enough to the average annotation score of a bread, they are rewarded with a boost in their credibility score, or penalized if their annotation score is opposite to that of the community score after a fair number of annotations. 

But where should I go if I am clueless? Here comes our Trending section.

At bRead, we believe that users should have a more natural choice to engage with the trending topics. Thus, we have found a new way to rank the posts that ensure diversity while ensuring that the breads receiving the most traction are at the top. This allows the user to engage in many topics and not limit them only to the most common ones.

🍞 User Study Methodology

We conducted in-depth semi-structured interviews with 5 users - 3 identifying as Male, and 2 identifying as Female. These users are pursuing a technical education in computer science, and are thus actively aware of technology and its impact. The interviews had a 2-fold aim. The first aim was a behavioral inquiry with regard to people’s usage of the bRead. The second aim was an attitudinal one, where we observed and probed people’s reactions and perceptions towards the possibilities of consuming content over bRead. We collected notes from interviews, our observations and user quotes to create this affinity diagram based on inductive pattern matching. With this, we unpacked insights about (1) the usage of the platform with emphasis on polarity adjustment, (2) people’s perception towards the platform affordances and concept and (3) people’s reaction to such a concept.

🍞 Insights from Actual Users

The reception to bRead’s concept of widening views was very well received.

A participant described why they would use it by saying

    "I just want to know what people think, and how do they think that".

A user who was interested in reading arguments defying climate change said that it was

    “Kinda like seeing the full picture... an opportunity to see what all is out there.”

A user who thinks of the platform as breaking the filter bubbles said

    “I think I really like this idea; I would probably use it for some topics that we talk about”.

Interestingly, a user saw this platform as a way to be informed about multiple topical opinions:

    "I wouldn't use it everyday, but once in a while to see what people think about".

Another user thought the platform can be used for quick public opinion surveying.

The polarity adjustment filter was also found to be intriguing.

    “The visual filtering of polarity looks really cool. To see the extreme content going away is interesting.”.

This visual difference worked to generate impact, as evident by a user said

    “On less intensity, probably less shouting and more discussion would be there”.

The going away of high polarity posts was appreciated both by the users, as well as the guest police inspector who came for the presentation. The latter expressed his excitement and thoughts about the potential utility of bRead to survey public opinion and help the police be informed so that they can take proactive measures.

That said, people are very guarded with their responses. For topics a user closely associated themselves with, they became annoyed or upset on seeing the other side’s strong views. For a high polarity setting, they dismiss or have negative perceptions about content being shown. However, for topics where a user has no standing, we didn't not see any negative feelings or thoughts even at high polarity. Here, they cherrypick information from both sides. This is extremely important, since bRead can benefit users who already do not have very strong views about a specific topic.

A user summarized their experience with bRead as

    “It's like twitter but so that I do not get caught in a bubble”.

They said

    “It’s good because I would know how to react if I met a climate change denier in real life. I can know what they are talking about and what their reasons are for it.”

Thus, we see bRead as a way to be informed about topical issues, as a way to understand (and sometimes dismiss) the conflicting views, and finally, as a way to understand and shape one’s opinions when exposed to a new topic.

Turns out, bRead has much more potential than we imagined! Try it out yourself!