- Stretching the facts, exaggerating claims are par for the course for startups
- Talented data geeks from UM come out top guns in online academia datathon
I had a glimpse into the power of big data analytics and machine learning algorithms on Thurs, May 3 when I was given a briefing into the results of an analysis The Center for Applied Data Science (CADS) did on the upcoming 14th General Elections (GE14) in Malaysia.
You can read the interesting article that CADS founder and CEO, Sharala Axyrd wrote where she shared some of the results but not all, and that’s because we realized there was an inherent bias in the analysis of their quick experiment that delved into tweets about GE14 from March 1st to April 30th. This was because the incumbent ruling party, Barisan Nasional, seemed to have given up on Twitter, likely because their power base is in the sub-urban and rural areas where folks are not into Twitter for their news or personal expressions. The other factor that distorted the results was language. CADS looked into tweets in English only.
To do a proper sentiment analysis of GE14 would require looking into tweets in all the four major languages of Malaysia ie Bahasa Malaysia, Mandarin, Tamil and of course English. But a proper analysis was not the intention of CADS. They just wanted to illustrate the power of data and analytics and machine learning algorithms in any major life shaping/nation building event.
However, if you are curious to see the results of their two month analysis, March to April, do reach out to Sharala and CADS.
Meanwhile, in Indonesia, which thankfully is not having an election, buying gold by retailers has gone digital thanks to EmasDigi. What’s interesting about this startup is that they first focused on B2B market but then moved to the retail side targeting students before recognising housewives now. The article also shows how startups like to stretch the truth.
For example EmasDigi claims that most housewives in Indonesia now use their smartphones to shop or carry out transactions online. They also claim that housewives find it interesting to invest in or trade gold via a mobile app. I would have been skeptical even if they had claimed the data was for Jakarta only.
But startups will be startups and stretching the facts, exaggerating claims are par for the course.
Meanwhile, it is no stretch of the imagination to say that Anthony Tan, cofounder of Grab, after having raised US$4.1 billion, now faces his biggest challenge – how to monetise the business model of Grab to justify the funding raised and keep his investors happy. You can read more about why he made it back as a repeat Digerati50.
Anthony Tan’s billion $ monetisation challenge
Meanwhile, hoping to make it in the future as DNA Digerati50 themselves (I hope), are a talented group of computer science undergrads from Universiti Malaya in Kuala Lumpur who, last Sunday, emerged as winners of a global online academia datathon. Coincidentally, CADS were instrumental in bringing the competition to the attention of the UM team and faculty.
My congratulations to the team that was hoping at best to be placed among the top 20% of teams.
As I sign off on this late edition of Week in Review (it should have been published yesterday) do look out for our analysis on Wed (yes, we know it is election day in Malaysia) of a recent global startup report and how Kuala Lumpur fared.
With that, I wish you a productive week ahead and to all Malaysians, do go out to vote and know that you have the power to decide the future of the country with your vote.
University Malaya team wins world’s first online academia datathon
TusStar powered Selangor Accelerator Programme launches with 30 startups
Has cashless finally arrived in Malaysia? Not really
Investing in gold with EmasDigi
Machine learning algorithms dive into GE14 battle on Twitter-verse
Swingvy raises Series A bridge round as it expands into insurtech
Biggest challenge for Malaysian based startups – lack of global connectedness
For more technology news and latest updates, follow us on Facebook, Twitter or LinkedIn