عربي

5 things to know when mixing data science with startups

Arabic

Click ‘CC’ for English subtitles.

Is it ethical to ask for phone location as part of your app permissions? Will you go out of business if you don’t feel savvy enough to “embrace data science”?

While ethical use of data may be the biggest issue related to data science on a personal level; on a company level yes, you will not last if you don’t start exploring the options provided by data measurement and analytics, says Microsoft Gulf data architect Rima Semaan.

Semaan was one of a number of prominent female MENA data scientists at the AUB Stanford Women in Data Science conference on Friday. It was supported by Stanford University and held in parallel with other events in 75 locations around the world, including Ramallah’s Birzeit University and in Qatar.

The one-day event covered data science in healthcare, Internet of Things (IOT), and its applications in business in MENA. Roundtables discussed the applications of data science in education, finance and aviation; data privacy and application of data science in IoT.

The big data and business analytics market is expected to grow by more than 50 percent between 2015 and 2019 and so are investments in the market in the same period.  

 Heavy presence of women during the event. (Image via AUB Stanford WiDS)

1. Data science can be applied in all industries.

In aviation, data science is used in areas such as  predictive maintenance where sensors monitor critical parts over time, and in flight delay prediction, drawing on data such as global weather conditions and social media. The banking sector is using data analytics for fraud detection and to  check whether someone is eligible for credit.

Semaan talked about a project in a shopping mall in the UAE where sensors are used to gather data to find out the paths most taken by visitors and, as a result, find the store locations that will receive the most visits.

Siroun Shaigian, cofounder of Kamkalima which makes education tools in Arabic, pointed out a “big gap in using data science in education”. She encouraged extra training for teachers on using technology, to make progress in this sector.

2. Data privacy is important to users.

Users are often hesitant to give companies access to data they consider very personal for fear of how they will use it later. They are also irritated by being required to give access to this data in return for using a service. For example, Careem and Uber require users to give them access to a lot of permissions when downloading their apps.  These concerns have contributed in the creation of services such as Signal, an encrypted messaging app, and Protonmail, an encrypted email provider.   

3. The importance of ethics is growing in data science.  

The ethical use of data is becoming a bigger issue as technology is used to target and demean minorities or other sections of society. Asmahan Zein, president of the Lebanese League for Women in Business told Wamda “while each person has to know where to draw the line, and protect the right of others; technology makes that need on a bigger scale.”   

4. You need to know  mathematics and computer science to work in data science.

“Working in data science requires technical skills more than it does soft skills; your [communication] skills for example will not get you anywhere without technical skills to support it,” said Dr. Ayse Basar Bener from Ryerson University in Canada. “It is important for those working in the field to have a strong academic background in mathematics and computer science.”

Daniela Colombo, an advanced analytics specialist at Microsoft EMEA, said passion and a willingness to go back to school would allow anyone to enter the data science world, even if they are from a different academic background.

5. You have to build your own platform.

Even if you can find ready to use algorithms for data analysis, “you have to build your own platform if you want it to work the way you want”, said Fadwa Mohanna, cofounder of Dubai company Markelligent. “The data you give to the machine can be flawed, that is why you have to build it from scratch.” In case you were missing statistic programming, or experience in the field you are working in, you have to “hire a team with all these skills,” to have a chance of success.

Thank you

Please check your email to confirm your subscription.