GE14: Sheer lucky coincidences or brilliant yet subtle social media strategising by PH? : Page 4 of 4

 

 

GE14: Sheer lucky coincidences or brilliant yet subtle social media strategising by PH? : Page 4 of 4

Data Analytics and Malaysia GE14 – Final Thoughts

While Malaysia leaps into uncharted waters via the strongest display of democratic right exercising in our history, we are also learning and (we hope) finally appreciating the role that data analytics has and could have played further in the different permutations of the GE14 outcome.

Do keep in mind that this study was done using a singular data source (Twitter events) constrained within the chosen data parameters for each rendition of the study. We utilised a Social Media Analytics framework to run predictive analysis through data collection and cleansing, sentiment analysis, online behavior analysis, topic modeling, influencers’ network detection and generating actionable insights. With what is just a granular fraction of the potential input that true Big Data Solutions can crunch and analyse, we could already pick out the election strategies of each of the coalition and the campaign stories that they would have leveraged on.

Now imagine the consolidation of the entire digital world and the digital footprint of the candidates from each coalition party and even the voters, notwithstanding the highly controversial discussion surrounding data privacy, ethics and election governance and integrity, Data Analytics has more than proven, in several instances of different national elections around the world, that it can provide insights and even recommend recalibration strategies while never missing a beat of the heart of constituencies and country.

With any study involving sampling of data, we understand that our insights generated from a small sampling of the Twitterverse would not be an accurate representation of the Malaysian voter population. We do acknowledge this sampling gap and have incorporated sufficient buffers in our analysis models to overcome any major result skews. However, our predictions and insights have proven accurate and even materialised as actual election strategies undertaken by the coalition parties, purely attributed to the digital footprint and utilisation pattern by the Malaysian voter public. With the assumption that the GS Coalition and their target voter demographic have effectively taken their campaigning and election engines to be predominantly centered offline, the digital social media space was open for domination by either the BN or PH Coalitions.

Remember that just because it is digital in nature, social media and the sentiment volatility that it can create or correct will not be void of emotions. If recent cases of social media bullying and online behavior patterns are any indication, there is in fact an increased tendency for people to over-react to certain stimulus / incidence acts, especially when they have the shroud of anonymity masking them. With Data Analytics automating the laborious process of data collection and crunching the data to understandable graphs and tables, the traditional roles of Campaign Managers and Social Media Manager will start to execute actual strategies that they know will impact voter sentiments and actually have the data to back their decision in near-real time. Taking an unbiased approach to data opens the door to unseen insights. It provides a unique perspective that might turn out to be far more important than you ever imagined.

Another key consideration that we from The Center of Applied Data Science would like to emphasise is that this Malaysia GE14 Sentiment Analysis study was not run as a pure analytics project. While the technical team were putting in long hours just to build, refine and monitor the analytics model in the backend, we still required our team of subject matter experts to bring out the true value behind crunching this volume of data at such speed. This is similar to our philosophy that the transformation to become a data driven organisation should not and cannot be run as an isolated IT / Analytics project.

In our study, we needed people who could identify patterns from social interaction data, visualisers that could assist in testing the numerous hypotheses that we have formed and tested over the study, even down to understanding the psychological implication of certain “trigger words” or “actions” on the average public’s sentiment. In our efforts to capitalise on the vast potential of the digital data world and constantly evolving analytical tools at hand, we will always need visionaries who are fearless in not just questioning the unknown world of tomorrow but actively take control and shape it for themselves and humanity.


Sharala Axyrd is the founder and CEO of The Center of Applied Data Science (CADS). She hopes this GE14 Sentiment Study will inspire and kickstart many more history-impacting analytics adoptions in the future.

Readers who need more insight into the report can contact CADS.

For more technology news and latest updates, follow us on Facebook, Twitter or LinkedIn

 
Keyword(s) :
 
Author Name :
 
Download Digerati50 2020-2021 PDF

Digerati50 2020-2021

Get and download a digital copy of Digerati50 2020-2021