Who are Citizen Data Scientists? Q&A with Hossam Elsemellawy | Coursera Community
Coursera Header
Q&A

Who are Citizen Data Scientists? Q&A with Hossam Elsemellawy

  • 15 February 2020
  • 12 replies
  • 808 views
Who are Citizen Data Scientists? Q&A with Hossam Elsemellawy
Userlevel 7
  • Community Manager
  • 1376 replies

Citizen data scientist is a new role that has appeared in many organizations in the last few years. Who are citizen data scientists and how can they change an entire organization’s culture? And what are the qualifications required to be a citizen data scientist? 

Post your questions about the citizen data scientist role below. Hossam will answer as many questions as he can between 23–29 February 2020.

About the Q&A Host

Hossam is a data scientist at Saudi Telecom Company (STC), the biggest and most profitable operator in the Middle East. He specializes in building churn models, propensity models, behavior segmentation, market experiments, sentimental analysis, path analysis and other business excellence tasks for reviewing campaign activities and proposing how to enhance performance using predictive analytics. Prior to STC, Hossam worked with SAS and Teradata partners, responsible for delivering data science projects to many Middle Eastern telecom operators, banks, and oil & gas companies and delivering statistics and machine learning courses using SAS technology.

Currently Hossam is supporting many aspiring data scientists by providing guidance, discussing business cases and helping them to link business cases with suitable analytical techniques. Hossam has a master's degree in ImmunoInformatics, scientific publications, a published book, and professional certifications.


12 replies

Badge

Who are citizen data scientists and how can they change an entire organization’s culture? 

Userlevel 2
Badge

Who are citizen data scientists and how can they change an entire organization’s culture? 

Hi Adabo,

Thanks for starting the thread, I will try to make my answer as complete and short as possible.

Citizen data scientist is a new role in any organization that is willing to move toward data driven decision making, normally each organization is doing too many tasks and not in all cases they know that there is a window for improvement, normally they raise requests for only cases where they see obvious issues that prevent them from achieving their targets, and once they are able to formulate specific requirements, they start to look for support from consultancy to address the problem, which normally takes long time for studying the case, looking for data, analyze the data then come up with the solution.

 

Citizen data scientists are taking different approach, they are already a part of the organization who share same targets and objectives, they are already familiar with the business, data and process, they are well oriented about the organization structure and stockholders for all cases, they are also aware about data science concepts and technologies, so normally their work is mainly done through initiatives that came from their awareness about organization processes and their different view to come up with new solution for known or unknown problems but they know that there is a big window for enhancements.

 

so their role in the organization exceeds the limitation for building machine learning models as peer request but they could address areas that will have much better results if we deployed advanced analytics and machine learning in its process, so they took the initiative and start to implement it them go into their experiments on a limited base to see its impact . as peer that those people need access to organization data and a window for making some experiments then ability to see current organization status and board main objectives to make sure that their work is running within main organization priorities 

Userlevel 2
Badge

Building citizen data  scientist team, requires not to start from scratch to bring a team totally from outside of organization, but requires to select a team couch who will select the team from inside and outside your organization who could help him to fulfill his task.

The team normally are not charged as peer their delivery but as peer their impact, so they should have a great awareness about their organization business and local and global market competition plus deep understanding for what could be achieved by AI and advanced analytics 

Userlevel 1
Badge

Hi Hosam!

I just heard about citizen data scientist. I am glad you are here. 

  • What are the qualifications required to be a citizen data scientist? 
  • How to start a new career path to be a citizen data scientist? 

Thank you.

Badge

Greetings Hossam, 

Could you please, clarify the difference between "Data scientist" and "Citizen Data Scientist"? What other roles does this term "Citizen" add to the known data scientist tasks and roles? 

Badge

Greetings Hossam, 

 

What are the real obstacles that prevent organizations to move forward to build the local Citizen data science team, even though as we  personally believe on the power this new concept can make having reading your valuable article,  

Userlevel 2
Badge

Hi Hosam!

I just heard about citizen data scientist. I am glad you are here. 

  • What are the qualifications required to be a citizen data scientist? 
  • How to start a new career path to be a citizen data scientist? 

Thank you.

As we saw from first post, that citizen data scientist should practice his/her role through initiatives that came from awareness about available data, problem description and analytics capabilities and requirements, 

 

so he/she should have the following skills and they all come with similar weight:

  1. Core Awareness about business domain
  2. Good SQL and data manipulation skills
  3. Deep understanding for statistical analysis and root cause analysis.
  4. Ability to arrange A/B testing and pre and post analysis
  5. Excellent in building machine learning models using several techniques especially white box techniques like regression and Decision tress
  6. Good understanding for descriptive data mining techniques like clustering and association rule
  7. good presentation and communication skills

As you see that some of these skills could be collected from training and rest should be collected from experience and market so, he/she should first work in collect statistics and machine learning knowledge from courses, then pay much attention to the criteria for selecting his/her first employer that they should working on giving him/her required business and domain data knowledge. 

Userlevel 2
Badge

Greetings Hossam, 

Could you please, clarify the difference between "Data scientist" and "Citizen Data Scientist"? What other roles does this term "Citizen" add to the known data scientist tasks and roles? 

Hi Yahya,

Thanks for your questions

 

Normally the difference came from understanding your organization business and data, everyone starts as data scientist who wait till get requirements from business about specific problem, then go into process for looking for relevant data then validate data till apply  analysis then come up with results then finally go to business with the outcomes for more discussions.

 

citizen data scientists are not waiting from case to be raised from business, although they may work on clear business requirements, but they could start too many initiatives from their core understanding for organization data and business, so he/she could take the initiative and come up with too many new suggestions for enhancements that may not be considered as an area for improvement. citizen data scientist are attending the meetings as a partner who support to find a solution not just for requirements collection and development.

Userlevel 2
Badge

Greetings Hossam, 

 

What are the real obstacles that prevent organizations to move forward to build the local Citizen data science team, even though as we  personally believe on the power this new concept can make having reading your valuable article,  

Hi Sami,

 

Normally the biggest obstacle is the time required to capture your organization DNA (business, data, process, who is doing what), this discovery for your organization analogy is not easy at all especially if it is huge organization, normally there is no unified process for this discovery unless your organization is doing this task as a part of orientation activities. Analytics doesn’t admit with team borders, churn models have impact on marketing, sales, customer experience. so your challenge will be how to reach to everyone to communicate your findings and make sure not to repeat and reuse any existing efforts in your organization especially for data collection and cleansing.

Badge

Hello Hossam,

I’m just starting to learn about data science and am very excited about your Q&A!

I’m curious about your day-to-day work. What tools/software do you use as a citizen data scientist? How do you collaborate with other people on your team or in your organization? What are some of the most satisfying or rewarding things about this work?

 

Thanks for your insights!

Userlevel 1
Badge +2

Hi nice to find your valuable inputs. I’ve some Qs:

Q1: Is It part of org. Innovation Strategy or Dept Initiative?

Q2: How to differentiate from Other Posts/Committees like Review, Feedback, coordination?

Q3: How will you differ with CDO (Chief Data Off) or ADOs (Assst Data Off)?

Q4: Is It overlapping or unique?

 

Userlevel 2
Badge

Hello Hossam,

I’m just starting to learn about data science and am very excited about your Q&A!

I’m curious about your day-to-day work. What tools/software do you use as a citizen data scientist? How do you collaborate with other people on your team or in your organization? What are some of the most satisfying or rewarding things about this work?

 

Thanks for your insights!

Hi Julia,

 

Thanks for your important question, when we talk about day-today work, so it is vary from meeting with your business user to understand your business problems, follow current business situation and  capture your top management targets and strategies, then the data sciences team themselves may prefer to make some internal discussions to share ideas and recommendations with each others before going into normal analysis process for extracting cases, problem related data, data transformation and cleansing, then apply analysis using whatever statistical or machine learning method till come up with the results and build the data story telling presentation. More meetings may be arranged with IT people regards data quality or open access to one of new data sources and final for model automation, story telling presentation also is passing some through internal discussions before sharing with business and top management for some refinements and some support by figures. then finally meeting with your business users for sharing analysis details, conclusions and recommendations and proposed actions.

 

so as you see, it is not business only task, not technical only and not statistics and machine learning only task, it is a mix that become a culture of work for you and organization by the time of working together in too many use cases and projects.

 

The most satisfying in my opinion is to see the impact of your work one the whole company results, when we size the impact of data driven decision making and compare it with non data driven activities and see the huge difference, that is creating the satisfaction to the data scientist, 

Reply