Global Research Letters

Social Media Research with Digital Methods

The subject matter for today is social media newspapers names, newspapers name research newspapers names, newspapers name with digital methods. My name is richard rogers professor of new media newspapers names, newspapers name and digital culture at the university of amsterdam. What i would like to do today is talk about a number of techniques newspapers names, newspapers name to study social media newspapers names, newspapers name platforms and these techniques newspapers names, newspapers name . Broadly known as digital methods are tool based but also research newspapers names, newspapers name question based so this. Is these um the examples that i’ll give you of the projects that i’m covering uh will give you a sense of the kinds of research newspapers names, newspapers name questions that are typically posed in digital methods work as well as some of the techniques newspapers names, newspapers name specific techniques newspapers names, newspapers name that are used and some of the outcomes as well so how productive these methods can be and in order to situate. What we’re going to be talking about today. I would like to think of social media newspapers names, newspapers name as being a social media newspapers names, newspapers name research. These days is in a is in a context a larger context of internet history and in particular the context of studying social media newspapers names, newspapers name as a as a comment space as a space of discussion. Uh as a space of of debate as a space mainly of leaving comments in your posts or making different kinds of contributions whether they’re authentic toxic sincere incidents here but generally speaking to think of social media newspapers names, newspapers name research newspapers names, newspapers name as research newspapers names, newspapers name about the conversations that we are having online. And i i want to give you um one of the early examples of using techniques newspapers names, newspapers name to sort of map and study the common space and this is the political blogosphere. This is work. That was done by a lot of damak and colleague to map the liberal and conservative blogospheres in the u.s and look in particular at their linking patterns so which blogs linked to which other blogs and and you see here a kind of classic story of polarization where in the red you have conservative blogs largely linking to one another in the blue the liberal blogs in the u.s context leading to one another and some shared linkages but the story is about the lack of shared linkages between them between the two spheres.

And when you look to this work more deeply you’ll see that the conversations are quite different. The language is quite different. These in some sense realities are quite different. Now um i want to talk about the contemporary study of social media newspapers names, newspapers name and and the various techniques newspapers names, newspapers name of studying uh each of these platforms. I’ll i’ll talk about twitter. Facebook google web search which people don’t normally group in in social media newspapers names, newspapers name research newspapers names, newspapers name but nevertheless instagram youtube. And then i’ll get a little bit into the deep vernacular web so reddit 4chan and some alternative uh tech platforms telegram and then the newcomer tick tock so. I wanted to start off with with twitter and give you a couple of examples right off the start of of pieces of work so this one here is of the urls that are in tweets by supporters of uh then candidate donald trump versus hillary clinton in the run-up to the 2016 elections. And you see here. The again a similar story to the one that we just showed with the political blogospheres where you have the hillary supporters in their tweets linking or referencing particular media newspapers names, newspapers name sources newspapers names, newspapers name and trump supporters in their tweets referencing particular media newspapers names, newspapers name sources newspapers names, newspapers name and very few shared sources newspapers names, newspapers name . Are referenced now if you look also a little bit more specifically at the sources newspapers names, newspapers name you see also on the trump supporter side the referencing of some quite extreme sources newspapers names, newspapers name versus on the hillary side which is a little bit more mainstream so when you begin to so not only is it a story about different referencing or perhaps people use the term echo chamber etc but you can also drill into the types of sources newspapers names, newspapers name being referenced. The next one is a technique newspapers names, newspapers name to segment an audience or map. A sort of set of what kind of movement if you will. This is the old right and these uh are um pictures of those who mention the core members of the alt-right and if so if you mention all eight of them you’re on the first map seven or six or or seven or six or now five and so you’re you’re beginning to show um this sort of audience for the um the alt-right so here it is again so there’s the core and then who mentions all eight of them there you have it and then who mentions uh at least seven and then who mentions at least six and and so forth so this is a that’s a that’s an audience uh segmentation uh technique newspapers names, newspapers name that you can you can undertake in order to sort of map and map it like an audience or a group formation or even an extent social sort of a social movement so those are two examples um so one is uh studying the shared sources newspapers names, newspapers name of supporters of particular candidates political candidates.

The second one as i just mentioned the segmented audiences you can also look at into retweet networks so who retweets whom. This work is often done to look at for example in political research. Um here it says parliamentarian so so because most parliamentarians in the western world all have twitter accounts and then you can begin to study um their their own networks uh and their i who who they retweet and who they don’t retweet and you can. Then look at diversity plurality or or again group formation. You can also study publics. The approach that i oftentimes put forward is to study competing ones competing hashtags and thereby potentially antagonistic public so um black lives matter versus blue lives matter etc just a couple of other ones. I’ll mention briefly. Their single hashtag newspapers names, newspapers name analysis has often been disparaged in the literature. However since me too. I think it’s been making a comeback but but arguably it’s one of the areas where it would work quite. Well is for summits so summits normally have political summits or other kinds of or olympics or they normally have a kind of dominant hashtag. Which if one were to query and make a tweet collection about that hashtag newspapers names, newspapers name one could explore that summit the tweet collections of public figures tweet collections of political leaders populist leaders all politicians in a particular election or other kinds of public figures oftentimes these days in particular.

There’s a lot of work being done. On artificial amplification the extent to which there are bots in a particular space. So if you were to make a tweet collection about a summit or about an election you could measure the bot activity and see the extent to which this bond activity is artificially amplifying one side versus another and then of course most recently. There’s been quite a lot of work on disinformation studies. So-called fake news and this has been facilitated in particular in twitter research newspapers names, newspapers name by the availability or twitters making available particular data sets that the two well-known ones are those of the so-called russian trolls and and and they think also is a set of iranian trolls. And they made this this. Uh this these data sets available that you can explore. What’s interesting about them. Is the kind of privacy policy. That’s built into them and the ethics thereof so twitter has the idea that if uh of a very specific definition of a public figure and that is if you have 5 000 or more followers you’re considered as such by twitter so in their data sets they hash or anonymize all users under under that figure of followers and those above. They don’t hash so that’s interesting to point out okay. I’d like to now move over to facebook. Facebook is the largest of these uh platforms. It’s also in some ways the most significant in this in the sense of what we’ve been talking about for for these sorts of disinformation studies etc. Also i’ve had the reputation recently of um being this the site of quite a lot of problems around uh elections in particular and this is a piece of work here that i think is kind of interesting to to point out was in the wall street journal it actually ran for quite a while they simulated what they’re called red feeds and blue field feeds. So if you were a conservative or like like let’s say a trump supporter versus in the previous election. Hillary supporter in the more recent election uh abide in supporter the kinds of sources newspapers names, newspapers name that you likely would encounter in your feed and um and then there and then and then the sort of uh the kinds of narratives about social issues or about the other that you would that you would be coming across more regularly and and then comparing these um uh is the is the task at hand.

So there’s the there’s the simulation um another technique newspapers names, newspapers name is something we call most engaged with content analysis. Uh and this is a technique newspapers names, newspapers name where you figure out which posts on facebook have received the most engagement and um so engagement is like shares comments or now reactions shares comments and then you add those up and and you see for example you take a set of pages of a particular a group or group formation. You see here in this example. It’s the alt light which is the sort of less extremist alternative right that movement and we took a number of their pages on facebook and we looked over a particular period of time. In this case i think it was a year and then looked at which posts received the most engagement across all these pages so in some sense. Your your your kind of demarcating a particular group or group formation and then within that over a period of time. What animates them. I should just also mention um before moving on that that the visualization here is a tree map so the the the uh post with the greatest uh amount of engagement uh is is is resize that is the largest size and we also um placed in this graphic also a sort of sub-categorization of different types of posts. So the sub-categorization was whether the posts were about white supremacy whether posts were about counter jihadism or islamophobia etc. And we in fact found that it was that it was the anti-muslim posts that animated this this this group the most. It should also be pointed out that um this work uh does not explicitly take into account or can even perhaps take into account uh content the effects of content moderation.

So it could be that posts about you know white ethno nationalism. Were the most engaged with but that they have been subsequently removed so the the the next one is um a network graph network visualization and here. Um it’s a visualization of pages that link to one another so on facebook pages can link to other pages. They can also um have have related pages so you could in fact make a network on the basis of facebook’s recommendation algorithm or make a network on the basis of uh which page links to which page and this again um is the alright. We were doing a lot of work on this at the time and use and you see that you can um link have a look at the intellect page network and then by cluster label them and then see certain sub cultures subcultural movements or sub movements within a larger. Uh a larger space. So so you see the alt-right with a bunch of sort of subcultures like like uh vaporwave or uh others white pride etc so for facebook um generally uh what you see are first the simulations that i showed you so these are these are simulations of um of news feeds. According to let’s say ideology the second one is that i show was the most engaged with content analysis. So you take a make a list of pages choose a time frame and then look across all pages all posts on all pages to see which ones we’re engaged with the most and then that answers the question of what animates a particular group you can also do inter liked page analysis. Um now the the most engaged content analysis. You can do these days. Uh still with face pager which connects still to the facebook pages api which for most tools and for many people was discontinued. But that one still continues or you can do this work manually. So yeah so. The same goes for inter like page analysis. Now facebook has rolled out after basically deprecating. The pages api rolled out its own tool originally built for marketeers crowdtangle. There’s also a marketing tool called buzzsumo that you could repurpose to do most engaged with content analysis but this type of content analysis or engagement analysis.

Let’s say is for web urls on facebook so how well do particular web urls do so rather than facebook posts. That’s that’s very very different but also could be interesting and finally facebook does have a an api political ads archive api which one can also look at and do some work with. There are also other projects to uh archive facebook. Um the one in particular that i’m referencing here is a counter archiving project but there are a few others where one collects a set of pages on a particular topic like like russian disinformation pages which are likely to be taken down and then one. Can you know have that archive keep you know so so. This is also a particular way of of trying to do facebook research. In the times of what’s referred to as platform lockdown so the deprecating of uh platform apis or other other ways in which facebook in particular but also instagram and others are making more difficult for researchers to get to get research newspapers names, newspapers name data. Okay i would like to move now to google web search and in particular. Uh talk about a few ways of looking into sort of. Let’s call them hierarchies of credibility um this it’s a web epistemological concern so you know which sources newspapers names, newspapers name of information have the privilege of providing users with it so with information so which sources newspapers names, newspapers name uh rise to the top and which ones uh don’t uh for particular queries that’s one and another one is is has been in interest recently is about so-called political bias of um of big tech or silicon valley tech. And that’s what this particular project looks into here. This is a project where one queries google. And when you query google you have to think about personalization so for this. We use a research newspapers names, newspapers name browser and we also use to choose the region setting so this was a project that was about the u.s which was region us and also used a u.s. Vpn so this is and we we logged out etc so this is all ways in which to kind of disentangle ourselves from the from the object of study so so personalization and especially geographical personalization uh doesn’t affect the results and so this is the results for queries three queries.

Um guns firearms one query. The other one is second amendments and the third one is gun control. And so here you’re looking at you’re looking to see what kind of sources newspapers names, newspapers name are returned for a quote-unquote sort of neutral or neutral-ish query uh versus a conservative query and a liberal in the u.s sense query and uh and what we found quite remarkably is for all queries. The top three four results were in themselves quite neutral. Uh and so there were wikipedia. Um things like this uh and then after that you had in the returns what you could call sort of left of center or liberal. So that from from news organizations to to ngos and then quite specifically conservative either news sources newspapers names, newspapers name or other types of sources newspapers names, newspapers name conservative ones were encountered at about result 35. So this is quite uh. Yeah it’s kind of quite radical findings or maybe not so um but it would be. I think to publish them as i think they would be of. Great interest to a particular side of that that debate these days. The second one that i wanted to talk about is what we refer to as um again source distance so how far from the top are particular sources newspapers names, newspapers name for certain queries. And what you see. Here is a visualization of what we were studying at times. Problematic information in particular uh election-related issue spaces online. So if you uh queried for certain politicians names and social issues where in the results were if anywhere. Uh were so-called problematic sources newspapers names, newspapers name and and this these could be anywhere from sort of extreme sources newspapers names, newspapers name to conspiracy ones or imposter news organizations etc. And so you see. The in. The visualization in red is where problematic sources newspapers names, newspapers name were encountered for particular queries and so for one or two queries. You can see there was quite a lot and so and you could you also get into here like the the the politics of problematic information because it was you know it’s in this particular example.

It’s a associated in particular with populist leaders and populist parties so for google web search uh more generally the the techniques newspapers names, newspapers name that i was referring to concern this quote-unquote source distance research. So how far from the top are either in the first instance sort of conservative versus liberal sources newspapers names, newspapers name and in the second instance problematic versus non-problematic sources newspapers names, newspapers name . But there’s also other work that people do and in particular they look into the extent to which google returns in the in the top first of all their own properties. So you know youtube videos google news etc um and so this is about you know google critique and preferred placement or even paid placement things like this and anti-trust concerns but also people look into the extent to which google can be or is being manipulated certainly around particular high profile events like the loss of vegas shootings for example where just after the shooting a 4chan post was towards the top of the returns in google. Quite infamously so i’d like to now move on to instagram. Instagram is a social media newspapers names, newspapers name platform. That in i think 2016 it was. It’s uh it uh saw its api. Uh shut off and so this was another example of a quote-unquote or an early example of quantum quad platform lockdown nevertheless. Since then. There’s been quite some work on building. Different kinds of scrapers. So you don’t get the quantities of data that you got in the past but you still get some so and this particular example here is again an example of this sort of source distance. Work so here you query uh instagram in two ways um you can you can query hashtags. Uh as well as geocoordinates and the combination is also very interesting so in this particular case first this is these are hashtag newspapers names, newspapers name queries these are queries for um again. This particular project that we were working on the politicians this was for a dutch ministry and we queried the hashtags of the politicians names as well as certain social issues and we looked at the extent to which the posts that were returned had uh sort of divisive content so what were the posts that had were returned and also had the highest engagement so this is then the top so those with the most engagement are at the top and and then you know and on down and then this explores ideas about the extent to which um you know particular type of sort of extreme or or problematic content is is the ones that are that are most circulated most engaged with or or people like to to click on etc so then the next example um is another type of thing that one can do and this these are follow follower ecologies so who follows whom you can make a network out of this and in this particular case uh again.

This was a political space and we found three distinctive clusters one of a kind of mainstream news cluster. The second one of a kind of mainstream uh political party uh cluster and then the third one is a kind of emerging sort of right-wing ecology. This was in a this concerned a political space or political instagram in the netherlands and then the third one is something that draws quite a lot of interest um both in like anthropological but also celebrity studies and and it’s the study of fake followers so and or or in the social media newspapers names, newspapers name company parlance inauthentic accounts or or inauthentic uh behavior. There was a quite a well-known article that was written in the new york times in 2018 about how celebrities and public figures from all different sectors. Or walks of life all were discovered to have um purchased or at least somehow acquired or somehow came upon quite large quantities of quote-unquote fake followers. So one can do this sort of work. Uh for you know a set of public figures in this particular case. It was a set of media newspapers names, newspapers name organizations as well as a set of politicians to look at the extent to which their follower counts have been artificially inflated and this is in some sense the study of symbolic power so there are a number of different uh means by which uh one can think of um the study of uh of symbolic power but this is this is how one can to be sort of more important than one is and then one can then use that of course this particular one here.

Um is a concerns. An interesting analysis put out by phillips and milner concerning how over the past few years. We’ve seen the rise of trolling and the rise of insincerity on social media newspapers names, newspapers name and we looked into the extent to which the posts around the u.s elections 2020 elections with hashtags related to the candidates at the time so it was trump biden as well as bernie sanders where the top posts the ones that were getting the most engagement were sort of insincere sort of jokey or ironic or or earnest so well-meaning that on the one hand and on the other hand divisive or non-divisive and what the particular research newspapers names, newspapers name project found was in fact that the most of the content of the political space was in fact earnest. Uh rather than insincere even when divisive so so. That’s what you the uranus is in blue and the ernest and then divisive. Well these are combinations as you can see on the legend so finally i mentioned at the outset that that you can also on query instagram for geocoordinates. I mean the combination is interesting so this particular project combination of querying geocoordinates as well as um hashtags or when querying hashtags look at the geocoordinates of the posts. So in this particular project this was on a project done in 2015 after the u.s supreme court ruled in favor of same-sex marriage there was the hashtag newspapers names, newspapers name that circulated right thereafter called love wins and that was a met with a counter hashtag newspapers names, newspapers name or anti-program called jesus wins. And so when you look at where these posts are coming from you could map in some sense a kind of geography of of hate. If you will so in sum uh we looked at um again this this sort of source distance exercise uh and then in this particular case with with most engaged with content.

So um how high up are particular types of content when they’re engaged with the most and and characterize that kind of content we looked into briefly follow-up follower ecology so who follows whom on instagram the study of fake followers. Uh there are a variety of tools and techniques newspapers names, newspapers name for this for fake followers studies. Uh the there’s there’s also the um the idea that that content uh that insincere content might circulate more than sincere or that divisive content might circulate more than non-divisive or that particular spaces are dominated by those kinds of types and finally geolocating publics and the idea of being able to kind of show a geography of of a particular a particular side taking on a on a social on a social issue. I mean you couldn’t call it sort of hashtag newspapers names, newspapers name publix or antagonistic hashtag newspapers names, newspapers name public public’s analysis so i want to now touch on youtube youtube for years has been quite generous in its data provision through. Its uh through its api and when when looking at the youtube api there are a few things that that you can do that. Stand out or when or just when looking at the interface so you can uh search uh youtube and uh analyze the sort of hierarchies and and the results in similar ways that you could study google web search for example and this is what you see here. Um this is a rank flow diagram a consistent query over a period of time. This is in fact the syrian war. I think it was the query and then which videos are returned per day. And what’s interesting about this particular piece of work. This is published as a work by bernard reader and colleagues. Is that you see that on particular days when a certain event in the syrian war takes place. The videos at the top of the search returns are actually a bit more extreme. And you could possibly. This is what one oftentimes refers to as the sort of excitability of of youtube’s algorithms.

So you can study um search engine returns so also on youtube you can create subscription network so channels can subscribe to other channels and you can see which channels subscribe to which ones and create a network from it also channels can feature other channels so in this particular work. This was again on the on the alt right and looking at the extent to which it’s kind of on youtube so these are kind of all right. These are sort of like internet personalities youtubers. Let’s say and seeing the extent to which they’re kind of well interlinked at least through subscribing to one another or not in fact they’re not not so well and then who features whom and so when we looked into this further what we found is that the feature networks actually show quite well business relationships so this is again so you can make larger uh networks of um of subscribers. I mean this particular one here. These are related channel networks. Uh that was a particular picture that was in the api and then it wasn’t in the api and then it would return and then it disappeared again but nevertheless it gives you a sense of the kind of networks you can get so so just like in facebook you can uh map um you know sort of the networks of of of the users so this can in this case the those the channels or you can um sort of create a network with the uh recommendations of of uh of the platform so so it’s like which channels are related to another according to according to youtube. There’s one other thing that i just wanted to show which i think is kind of interesting so recently a lot of the platforms youtube. Uh certainly have been deleting quite a lot of content uh this is called d platforming uh or um or there are other terms as well and the question is well. What’s being removed. And and what are the you know. What are the implications of that uh. For um basically the the study of public discourse debate censorship uh cancel culture et cetera. So in this particular case um researchers using 4chan which we’ll come to in a second made a list of youtube urls quite extreme uh content and then looked over time whether or not they’re still available and so on one dates you see.

That’s now you see uh some content on the next day. That’s there there you see that they’ve been deleted and so then you can look you can begin to look into the type of contents but delete the type that’s been left up. You can look into any sort of traces of um sort of you know whether this deletion was was sort of automated whether it was all done exactly at the same time things like this so you basically begin to probe um the the larger issue of what’s called the the politics of deletion so for youtube. Um you can uh study a number of things with the api. I mean i started off with with search um and i also talked about doing making networks. So who’s you know. Channel subscription network. So which shall subscribe to which and which channels feature other channels um also talked about content removal but one of the other things that is of interest and you can also do with with some top software developed digital methods is um look into the sort of recommendation algorithm and in particular what you could call the carousel or the up next. And there have been a lot of sort of research newspapers names, newspapers name into this journalistic as well as some big data work on the extent to which the algorithm is designed to keep you watching to to binge watch and and and that’s done through providing more and more extreme uh videos and so you could in fact look into that uh and you could also do that comparatively you know compare a number of different different spaces so you could. You can start with one video that’s like a like a you know. A news broadcast where experts appear and then another one with sort of youtubers internet celebrities mark stream figures and then compare the recommendations to those videos one after the other and to see whether they both end up in extreme spaces or only one. Does i want to now talk about reddit and move now into the deep vernacular web term that was coined in amsterdam at the open.

Intelligence lab and this is a term that refers to sort of spaces where the users are not public facing but rather are anonymous and where there are in some ways a kind of different sorts of sorts of subconscious. Now now reddit in reddit for one uh was known for a while as having quite um quite some quite sort of extreme subcultures i mean it also has very somewhat quite political stuff but also um very innocuous stuff these are all subreddits so the the major one uh that was being studied quite a lot especially during the trump uh period was the subreddit called the donald also the largest it was it was thought to be the the site for um what was referred to as mean magic and sort of memeing trump into the into the presidency. And so so all of reddit is archived. Um and you can query that uh that archive. It’s a it’s a push shift archive and we also have created tools ourselves that build on top of that so one can look at for example and this is what you see before you a number of subreddits in a particular language space so um so this is then this has been dutch reddit and then all of the subreddits there and then those that have uh problematic information so one can sort of basically study per subreddit the language use of the activity etc but one can also study um sort of like national national reddits if you will. So that’s um so so. The single uh subreddit analysis um uh or and this also one can study the sort of traveling of ideas from one particular part of the deep vernacular web for example from 4chan to reddit and then off to sort of facebook and or twitter etc. This is the sort of what is referred to by the open. Intelligence lab as a sort of normification normification approach now 4chan. Um is is quite notorious. It’s the site of of quite quite extreme content. Um but it’s also the the side of has very specific subculture. Kind of kind of trolling.

Kind of jokey sort of edgy. Also a lot of a lot of bad language etc especially on its most or its biggest board poll so like with reddit studying subreddits here. You can study individual boards. So we’ve archived a few of these and the outputs of these can be for example. These are image boards so the outputs can be image walls. So which kinds of images are being circulated at particular points in time and you can use software like like image sorter image plot to make image walls and then group the images according to formal properties or show them over time. Uh just to see the the changes in that particular in that particular space. A word of warning. A lot of the imagery is is quite quite offensive. Although you’re also i mean this is also the place. Where a lot of um sort of edgy and leading edge. Uh memes are being developed and circulated a lot of them uh are you know racist and symmetric et cetera. So this is this is something else to talk about. But but as with uh reddit you can study individual boards or you do cross board studies or you can begin to think of 4chan as having sort of national spaces so whilst users are anonymous they oftentimes will use country flags so you can group posts by country flag and explore for example. The sort of you know extremeness or innocuousness of reddit in a particular particular country or a particular language space. I also wanted to show you um this is uh this is the uh the ephemerality of of posts or general posts in particular. I read it so so after a number of contributions per what’s called thread the the it automatically gets deleted however the discussion uh oftentimes continues and and and this is a way of thinking about studying. 4chan is is in some ways what you could think of as the sort of continuity despite ephemerality so what what continues there despite the the continual deletion of these of these threads so so as i said single or multiple boards or you could delineate a national space and then you can also study on these these threads telegram telegram recently.

Well a few years ago was associated. Uh with isis content and one of the reasons is it has a reputation for being really highly secure uh and and anonymous and a very. Let’s call it. Liberal content moderation policy since then it’s cracked down on that kind of material but a lot of the extremists or extreme internet celebrities that were deep platformed from mainstream platforms like youtube twitter. Facebook instagram have fled so to speak or migrated to telegram into an alternative tech ecology that includes telegram on telegram. What you can do is you can form a group or you you or you can um uh set up a account where other people can subscribe to you and in some ways follow you. So this is how a lot of the sort of extreme internet celebrities have been using telegram and uh and what we did uh when we looked into a number of them. I think we took about 20 of these folks who were de-platformed from the mainstream platforms and then we. We looked at their content. And what they were linking to which and in particular which other platforms they were linking to. So this is a link analysis of of uh of a set of uh internet extreme internet celebrities and and what we found uh in particular was that there are certain platforms that they still link to um that they’ve been thrown off of like like twitter and youtube so um but then others that they’ve also been thrown off of they hardly link to it all like facebook and instagram. So what we found is that that facebook and instagram are well first of all. We found that we could do cross-platform analysis by just looking at this this single platform but also that there are some platforms are still relevant to them. Despite being de-platformed whereas others are not relevant this was in some ways. Try a partial answer to the question of for whom um is does deep platforming work and it seemed to work for facebook and instagram in the sense that they are no longer of interest to these extreme internet celebrities.

What’s also interesting here. Is that the one space that is being kind of rejuvenated by them. In some sense through through links is the web um so that so social media newspapers names, newspapers name of course has followed the web um a lot of portions of the web are far less healthy these days because many have fled the web to social media newspapers names, newspapers name but kind of ironically if you will uh the web gets a gets a bit of a comeback here from from these extreme internet celebrities. Another thing that is of interest is to look at this larger alt tech ecology so this well when looking and then looking at sort of telegrams place in it but also other alt uh tech platforms place in it so we took all of the these uh extreme internet celebrities we looked at which platforms they. They still are a part of uh where they still have accounts and and then also the alternative ones where they have new accounts and made this kind of map of um and and did find indeed. That telegram is quite central although not as central as it were just mentioned as as their own personal websites making a comeback but also some of the other alternatives like bit. Shoot the alternative to youtube so with with telegram um we studied the public channels so the channels that are um used to try to get a follower base or a fan base you can also study public groups and you can study the place of one of these alternative platforms so in this case telegram in a larger alt-tech ecology i want to now conclude with tick-tock so tick-tock is uh interesting. Because it is not only um i mean one can study you. Know the youth culture the sort of the the and the trends of creative expression on this medium but also it’s also um increasingly a political space so just in the run-up to the 2020 u.s presidential elections tick tock posted. This notice saying that it wouldn’t stand for inauthentic or or extreme content and also asked others not to post it and what we then did is we looked into a number of political hashtags.

Uh around the three candidates at the time um so trump biden and sanders. The ones that were still left around i think was super tuesday somewhere at the end of march early april uh 2020 and also specific political hashtags. So there’s this maga one and looked into the the kinds of um creative expression so how we’re so on tick tock. There are certain trends like there will be a certain uh noise or audio clip that will suddenly become trendy and everyone will use and and they also signify certain things so there’s one which is used to call something into question so what we what we found is that a lot of the special effects being used in the political space. Were to call into question. Uh mainstream media newspapers names, newspapers name accounts of certain things so it becomes a kind of misinformation space in a different way than one normally thinks of misinformation as sources newspapers names, newspapers name or stories. This just sews doubt we also looked into. This is using a tick tock scraper. We were scraping. Um the most engaged with videos per candidate query as well as some some other hashtags what was interesting. Incidentally is that hashtags. Don’t demarcate political spaces as as you might think they would so um both videos that you could say on the left and on the right use all um the political hashtags at the same time so hashtags are used for audience generation as opposed to for that’s called identity politics so in this particular case we then looked into the the extent to which sort of problematic videos were amongst the most engaged with content. Uh in um on what’s called political political tech talk so for tick-tocking one can study um again forms of creative expression and then within particular hashtag newspapers names, newspapers name spaces and what we were looking at in fact was in some sense um public so those using but but that term i think does not really work so well in in tik tok but nevertheless you could think of it that way. Because they’re they’re you know not. All hashtag newspapers names, newspapers name uses is multi hashtag newspapers names, newspapers name per video and you could certainly see how um uh sets of videos or you know or on one side of political spectrum and on the other side of the spectrum you can also look into top content so again this is.

This is a kind of source distance. So so how far from the top is uh. What have you so in this particular case we looked into misinformation we also discovered. Incidentally that tick tock with the with its with its sarcasm especially the use of audio certain kinds of sounds um to be sarcastic. That was the way in which uh mistrust or even misinformation was sewn through satirical sound. So what i’ve done is introduce to you. The study of of different kinds of techniques newspapers names, newspapers name per platform to study the comment space or social media newspapers names, newspapers name more more generally and in particular looking into certain approaches. That are quite quite dominant or techniques newspapers names, newspapers name . That are quite dominant. When you in some ways follow the fields available to you in the apis or or the or the dominant features of of particular platforms and then unpack those so studying recommendations. So that’s on the. Let’s say on the algorithmic side or engagement. That’s on the user side so thank you very much.

Where to find great research papers?

Various great research journals such as Global Research Letters are a great option and way to help you look up impactful research papers with a great format. Here, you will find a number of various research papers that are provided and made available to you in the journal, which will help you write your own paper.

You can very easily find papers on a variety of topics at Global Research Letters, which will help you with your own research work and understanding of writing and publishing research papers properly. With access to so many amazing research papers, you can practice and learn the process of writing research papers and their importance.

Leave a Comment

Your email address will not be published. Required fields are marked *