Politweet.Org

Observing Malaysian Social Media

Archive for the ‘Visualisations’ Category

Emergence of Dedicated Spamming Apps in Malaysian Politics

In previous articles published on our blog we have described various spamming strategies used in Malaysian politics:

Some of these strategies are still in use today. Some key points about spam:

  • Spamming involves repeating the same tweet or retweeting another user’s tweet across one or more accounts within a certain time period.
  • Spamming is a violation of Twitter’s own rules for users (that you can read at https://support.twitter.com/articles/18311-the-twitter-rules ). They list various factors that determine spamming behaviour one of which is, “If you post duplicate content over multiple accounts or multiple duplicate updates on one account”. ‘Updates’ refer to tweets and retweets.
  • Spammer accounts are identifiable via their behaviour on Twitter e.g. follower/following relationships; timeline content; tweet timestamp patterns. Tweet frequency, repetition and collaborative behaviour are the main traits we look for.
  • Twitter accounts are being used for personal use and spamming. This lends the appearance of a normal human Twitter user for anyone looking at their timeline and a denial when accused of sending spam.
  • Some Twitter accounts are dedicated to spamming and have no personal messages of any kind.
  • Some Twitter accounts have a block of spam in their timeline but otherwise appear normal. This means they only spammed tweets briefly. It is possible their login credentials were being used without their knowledge.
  • Spamming does not always involve automation via applications. Humans using mobile devices to repeatedly send identical messages across multiple accounts can still be identified.
  • It takes a computerised system to analyse and identify these users and their tweets and categorise spam.

Developing spam detection systems is necessary for us due to the frequent use of spam in Malaysian politics. Spammers who retweet other tweets are problematic for 4 main reasons:

  • They increase the retweet counter for the tweet, making people believe that tweet was popular
  • By retweeting instead of tweeting, users who are searching on a keyword or hashtag won’t see the spamming accounts. This means people who use Twitter won’t discover this activity unless they happen to find the spammer in the list of recent retweeting users for the tweet.
  • The retweet counter is not guaranteed to decrease if the spammer is suspended or deleted.
  • It is harder to prove to non-technical users that the account is a spammer. Direct links to the spammer’s tweet will redirect to the tweet that they retweeted, which is a detail they may overlook. The best way to show evidence to the public is for them to visit the spammer’s timeline and judge for themselves.

All timestamps used in this article have been adjusted for the UTC+8 time zone.

Read the rest of this entry »

Written by politweet

June 1, 2015 at 12:51 pm

How Twitter Users Reacted to News of Lee Kuan Yew’s Passing

Singapore’s founding father and first Prime Minister, Mr.Lee Kuan Yew passed away at 3.18 AM on March 23rd 2015. What follows are statistics and a selection of popular tweets about Lee Kuan Yew shared by users online in response to the news. Tweets from news agencies have been filtered out of the list. For a list of business services that we provide please visit our main site.
Read the rest of this entry »

Written by politweet

March 24, 2015 at 4:38 am

The Rural-Urban Divide in Malaysia’s General Election

108 out of 133 seats won by Barisan Nasional (BN) came from rural seats. 72 out of 89 seats won by Pakatan Rakyat (PR) came from urban and semi-urban seats.

While it is true that PR won every Chinese-majority seat, there are only 30 Chinese-majority seats in the country. That leaves at least 59 seats won with the support of other races.

When comparing Malay-majority seats, PR won more seats than BN in both semi-urban and urban categories. A Malay-majority seat cannot be seen as a guaranteed victory for BN.

In terms of the popular vote, BN obtained 57% of the popular vote in rural seats, 47% of the popular vote in semi-urban seats, and 36% of the popular vote in urban seats. Looking at the winning majorities of individual seats, the probability of BN regaining urban seats is low. This gap in the popular vote is illustrated in the infographics at the end of this post.

That is the picture of the political urban-rural divide. BN represents the rural majority and can retain power with rural and semi-urban seats alone. This election highlighted PR’s weak areas which are rural seats, Bumiputra Sabah majority and Bumiputra Sarawak majority seats.

Read the rest of this entry »

Written by politweet

May 21, 2013 at 3:00 pm

Partisanship of Politicians’ Followers by Affiliation

View the high-resolution image at (http://www.flickr.com/photos/politweet/7842715014/in/photostream/)

This network graph shows the connections between Barisan Nasional (BN) and Pakatan Rakyat (PR) with their followers on Twitter. It visualises the size of the followers exclusive to each coalition, and the size of the overlap. This version is colour-saturated to show the partisanship a bit more clearly.

This graph is based on a snapshot of followers taken on August 2nd 2012. There are 1,036,932 total followers.

How to read this graph

  1. The red node represents PR. The blue node represents BN. The nodes are sized based on the number of followers.
  2. Each grey node represents 1 Twitter user. A node is connected to BN/PR if that user follows BN/PR.
  3. Connections to BN are blue, connections to PR are red. Overlapping (purple) connections indicate non-exclusive users.
  4. Connection strength is based on the number of politicians followed by the user.
  5. Nodes that are connected only to BN or PR are exclusive to that coalition.

Facts and Observations

  1. PR has 314,302 followers. BN has 876,138 followers. BN has 2.788 times more followers than PR.
  2. Active followers have used Twitter in the last 1-2 months.PR has 188,045 (59.83%) active followers. BN has 507,899 (57.97%) active followers.
  3. PR has 151,822 (48.3%) exclusive followers. BN has 719,341 (82.1%) exclusive followers.
  4. PR has a relatively small group of exclusive followers compared to BN. The PR-exclusive followers are the left-most semi-circular clusters (forming a red ‘fan’), representing 48% of all PR followers. The remainder of PR’s followers follow both BN and PR.
  5. The BN-exclusive group are concentrated in the top and bottom clusters. Its clear that either cluster is bigger than the PR-exclusive cluster.
  6. PR-exclusive followers start following both BN and PR at an average rate of 1935 users/month.
  7. BN-exclusive followers start following both BN and PR at an average rate of 929 users/month.
  8. PR-exclusive and BN-exclusive followers switch parties at very small rates, between 19-59 users/month.
  9. PR has grown from 269,473 followers in March 2012 to 314,302 followers in August 2012. That is a growth rate of 16.64%.
  10. BN has grown from 624,227 followers in March 2012 to 876,138 followers in August 2012. That is a growth rate of 40.63%.

Conclusion

PR has a slow follower growth rate and a higher migration rate of exclusive followers to non-exclusive status. This indicates that PR followers are more open to receiving messages from BN.

BN has a high follower growth rate and a lower migration rate of exclusive followers to non-exclusive status. This indicates that BN followers are less open to receiving messages from PR.

Barisan Nasional is doing quite well for itself. One area that could use improvement is the number of active followers.

Pakatan Rakyat faces a number of challenges. First is to ensure their followers don’t move towards being BN-exclusive. Second challenge is to get more BN followers to move away from being BN-exclusive to being non-exclusive; or PR-exclusive. This will help expose users to PR policies and ideologies. If they can repeat the PR message to BN-exclusive users, that will help improve the migration rate and encourage even more users to follow.

Both coalitions need to acquire more active followers who will help spread their message through retweets.

Data
Breakdown of BN followers:

Total: 876,138
Suspended: 21,091
Observer (0 tweets, 0 followers): 123,825
Inactive (no change in last 2 months): 223,323
Active: 507,899

Breakdown of PR followers:

Total: 314,302
Suspended: 8,055
Observer (0 tweets, 0 followers): 25,398
Inactive (no change in last 2 months): 92,804
Active: 188,045

Observer accounts are potential dummies. However many people do use Twitter as a news reader, so you can expect a significant number of Observer users to be real people. Many users also leave their accounts idle until there is a live event happening.

Update 23/8/2012 (in response to comments on fake accounts)
The issue of a dummy account needs to be raised together with a definition of what is considered a dummy. BN’s high follower count has being attributed to fake accounts, bot accounts and inactive accounts by civil servants and UMNO members forced to sign up on Twitter. These are the types of accounts that would eventually be detected and fall under the Observer and Inactive categories stated above.

Many of BN’s followers can be attributed to the Prime Minister (@NajibRazak). Without Najib, BN’s follower stats are quite different.

Breakdown of BN followers (not including @NajibRazak):

Total: 374,563
Suspended: 11,400
Observer (0 tweets, 0 followers): 49,167
Inactive (no change in last 2 months): 60,243
Active: 253,753

446,146 (56.87%) of Najib’s 784,477 followers are active. Najib gets retweeted often by a large number of users, so the influence of the account during the next GE cannot be discounted. A detailed analysis of Najib’s followers is definitely something worth exploring in future.

Dummy accounts make it look like the social media companies in charge are hitting their target, so they can earn their commission. But creating 100+K dummy accounts does not seem practical. You need an email for each, then signup on Twitter, then click the verification email, login and follow Najib – it is a lot of work.

Written by politweet

August 23, 2012 at 4:22 am

A Network Analysis of Users Mentioning Malaysian Politicians on Twitter, April 2010 – November 2011

This visualisation displays the distribution of users who mention BN and PR politicians, based on the number of mentions they made.

1.Users are represented by grey nodes.
2.There is a blue node on the left, representing BN politicians
3.There is a red node on the right, representing PR politicians
4.Users are pulled to either node based on the number of mentions they made for that coalition
5.The blue-shaded area denotes users who are exclusive to BN and do not mention PR politicians
6.The red-shaded area denotes users who are exclusive to PR and do not mention BN politicians

There are a total of 72,214 users that made 1,002,027 tweets mentioning politicians. Read the full report below:

Written by politweet

December 15, 2011 at 3:38 am

Who tweets about race?


Who tweets about race? Not that many apparently, if you don’t count our politicians.This diagram is not intended to highlight racism. Just the mention of ‘Malay’ doesn’t mean the user is saying something bad about Malays, or saying that Malays are superior. But it is worth noting that out of the 16,600 users who tweeted to politicians (the Tweeple), only 9.9% mentioned race.So out of the 213,282 tweets by Tweeple this year, which includes retweets of what politicians wrote, only 5,793 (2.72 %) mentioned race. It implies that any talk of race by politicians whether positive or negative hasn’t gained much traction among the people, and that the people themselves don’t often bring up race as an issue.

The following are noticeable patterns in the ‘race mentioning tweets’. The frequency in brackets indicates how many users followed that pattern.

Pakatan Rakyat

1. [majority] Highlight different races at their events, or the race of people they are meeting. For example a Chinese YB tweeting that ‘many Indians came’ for their ceramah, or ‘having tea with Chinese supporters’, or ‘had good conversation with Malay taxi driver’.

2. [majority] Proactively mention malay/chinese/indian villages that have received aid; how their parties have a mix of different races;the support their parties have from the different races.

3. [majority] Hightlight the poverty faced by different races, primarily Malay

4. [half] Often defend themselves against accusations of racial favoritism. They commonly cite racial breakdown of statistics as proof. Sometimes they proactively tweet stats to show how much help (scholarships/contracts/etc) given to each race.

5. [half] Highlight racial remarks by BN politicians, e.g. ‘..that they don’t need Chinese and Indian votes’

6. [very few] Questioning the race of some BN politicians, whether ‘really Malay, or mixed’ or ‘bukan Melayu tulen’. Also proclaiming their own purity.

7. [very few] Demanding for posts to be filled by certain races, and accusing BN of allocating a high budget to public services due to Malay dominance in the area.

Barisan Nasional

1. [half] Question racial breakdown of statistics cited by Pakatan. Sometimes they raise issues e.g. how many scholarships given to each race.

2. [half] Question DAP’s positioning of itself as multi-racial when their membership is largely Chinese. Also criticism of DAP for allegedly letting go of their principles to appease PAS.

3. [half] State they will be there for their respective races to champion their issues, but will help other Malaysians as well. As one UMNO politician put it, “..parti berasaskan kaum, tapi bukan berasaskan rasis”.

4. [few] Highlight aid given to indian/chinese/malay communities, but much less frequently compared to Pakatan. Also the tweets are on more general,large-scale terms, e.g. ‘a community’ instead of Pakatan’s ‘a family’, ‘a voter’ or ‘a village’.

5. [very few] Highlight different races at their events, or the race of people they are meeting.

Both Barisan Nasional + Pakatan Rakyat

1. [majority] Divide issues based on race

2. [majority] Criticise PERKASA and Ibrahim Ali for bringing up racial issues. Not one supported PERKASA.

3. [majority] Stereotype voter’s way of thinking based on race.

4. [majority] Label areas based on race.

5. [half] Sometimes tweet about Ketuanan Melayu vs. Ketuanan Rakyat.

6. [few] Discuss Chinese school land allocation/funding/education quality

7. [few] Criticise parties on the opposing side in racial ways, e.g. “What has MCA done for Chinese?”, “UMNO is to defend Malays, but sells Malay land”, “If DAP had Malay ADUN it would get MB seat in Perak”, “Melayu hilang suara selagi UMNO berkuasa”.

8. [few] Malay unity and malay division due to UMNO/PR

Tweeple

A detailed analysis of 5793 tweets would be too time-consuming, so here are the main themes of the tweets:

1. Criticism of politicians for highlighting the race of people when talking about issues, e.g. ‘most Malays are poor’ and choice of candidates in by-elections e.g. ‘non-Indian in Hulu Selangor’. Majority of criticism directed at BN and DAP.

2. Retweets of whatever politicians said.

3. Racial slurs levied against politicians, political parties and other Tweeple. This includes accusations of racism.

4. The Malaysian First/Malay First etc. debate – comments and challenges to politicians

5. Criticism of PERKASA and Ibrahim Ali, either directed at UMNO politicians or retweeted from/to Pakatan politicians

6. Malay unity and division, and how politicians’ actions were affecting it.

7. By-elections always brought up discussion on how much support BN/PR has from each race in various areas and the activities there.

Most of the other tweets varied based on whatever issue was current. These are so varied it would be impossible to list them all. A couple of examples:

1. Denial that gambling is Chinese or Indian culture (during the sports betting license issue)

2. Criticism of PAS on some Malay-centric ideas, such as requiring Baju Melayu for civil servants on certain days of the week

*the majority of tweets related to food e.g. Indian food, Chinese restoran etc. were not included in this study.

*talk of Chinese schools and education has been included, and will be elaborated on in a future diagram.

*some associations and locations have race as part of their name. E.g. KL-Selangor Chinese Assembly Hall, Kg Cina Sijantung, Msian Indian Restaurant Owners Assoc. These tweets were not numerous but would be difficult to filter out, so they were left in. Some changes to the database are being planned to make filtering easier in future.

Written by politweet

January 3, 2011 at 9:05 pm

Posted in Analyses, Visualisations

Tagged with , ,