DataTalks #14: Israeli Elections - Polls, Election Data and Future Ideas

Our 14th meetup is hosted by Oracle Data Cloud and will provide an opportunity to have a glimpse at two fascinating projects which examine past data from Israeli elections!
These projects will serve as an analytical basis for discussing intriguing ways to examine the data that will be collected in the upcoming elections.

𝐃𝐚𝐭𝐞 𝐚𝐧𝐝 𝐭𝐢𝐦𝐞: Tuesday, September 10th, 18:00 - 20:30
𝐋𝐨𝐜𝐚𝐭𝐢𝐨𝐧: Oracle Data Cloud, Derech Menachem Begin 132, Azrieli round tower, 21st floor. Tel Aviv-Yafo

***********************
In addition, we plan to hold another meeting a few days after the election day for a “hands-on” session, in which we will combine forces and analyze the data from the official polling stations (location and time will be announced soon).
***********************

𝐀𝐠𝐞𝐧𝐝𝐚
18:00 - Gathering, snacks & mingling
18:30 - First talk:
Yonatan Lazar-Telem (“AutoPoller”) - How I created and analyzed my transparent poll.
19:15 - Second talk:
Itamar Mushkin - Election results from 1996 to 2019: estimating transition of votes between parties
20:00 - Discussion and brainstorming regarding the upcoming election day

Talks are in Hebrew and will be filmed and uploaded to the DataHack YouTube channel.

---------------------------------------------------------
How I created and analyzed my transparent poll - Yonatan Lazar-Telem

The results of the “Transparent Poll” and its complete data, which was gathered with the help of crowdfunding, were published in April 2019. This data provides a glimpse into the hidden aspects of the polling industry. In this talk, I will present the findings and the questions that arise with regard to the polls that are published in the mainstream media.

Before the meeting, you are invited to review the “Transparent Poll” data using this link: http://bit.ly/transpolldata

Bio: Yonatan Lazar-Telem is a data scientist at Mobileye and holds an M.Sc in Statistics from the Hebrew University. Founder of the "AutoPoller" project, which encourages a critical reading of surveys.

Relevant links:
https://shkifut.info/2019/03/transparentpolls/
https://www.facebook.com/AutoPoller/

---------------------------------------------------------
Elections results from 1996 to 2019: estimating transition of votes between parties - Itamar Mushkin

We can learn from each parliamentary elections cycle which party gained or lost power (in terms of seats in parliament), but we cannot directly observe transitions between parties. Nevertheless, using the complete voting data gathered in each cycle, we can construct a simple model to answer this and other questions.
In this talk, I will present the underlying model for the transition of votes between parties, its estimations, and the resulting insights. If time permits, I will present additional results from a data-driven analysis of the election outcomes, such as political clusters that emerge from the data.

Before the meeting, you are invited to review the elections dataset and kernels using this link: https://www.kaggle.com/itamarmushkin/israeli-elections-2015-2013

Bio: Itamar Muskhin is a machine learning developer at Precognize, and formerly an Operations Research Analyst at the IAF. He holds a B.Sc and an M.Sc in Physics from the Hebrew University. Likes wrangling data and talking politics, and finds it very time-saving to do both at the same time.

Relevant links:
https://www.themarker.com/techblogs/ormigoldstein/BLOG-1.7141220
https://www.themarker.com/techblogs/ormigoldstein/BLOG-1.6567019