Confirmed users
1,262
edits
All Hands/2016 Hawaii/electivesubmissions: Difference between revisions
Jump to navigation
Jump to search
All Hands/2016 Hawaii/electivesubmissions (view source)
Revision as of 21:35, 21 October 2016
, 21 October 2016→Products
| Line 146: | Line 146: | ||
This elective will be interesting for those curious to learn about frameworks for doing fast iteration projects.</BLOCKQUOTE> | This elective will be interesting for those curious to learn about frameworks for doing fast iteration projects.</BLOCKQUOTE> | ||
===Products 5: Design Principles for Connected Devices Projects=== | |||
* Presenters: Connected Devices UX team | |||
* Description: We'll review the Design Principles for Connected Devices projects, followed by Q&A. Bring your CD project, and we'll work with you to apply some of the principles! | |||
===Products 6: How a Non-Profit "profits"=== | |||
* Presenters: [https://mozillians.org/en-US/u/mconnor/ Mike Connor] and [https://mozillians.org/en-US/u/schen/ Susan Chen] | |||
* Description: Exploration on potential new revenue streams and partnerships | |||
===Products 7: Data Analysis and Visualization 101=== | |||
* Presenter: [https://mozillians.org/en-US/u/alessio.placitelli/ Alessio Placitelli] | |||
* Description: We will analyse a sample Telemetry dataset, going through the basics of data analysis with Pandas & Numpy on our Spark/Jupyter infrastructure. Feeling stuck with the data visualization capabilities of Jupyter? We've got you covered. A big part of the session will show how to visualize the data using matplotlib, plotly and others. | |||
===Products 8: Defining Firefox Product Strategy=== | |||
* Presenter: [https://mozillians.org/en-US/u/pdolanjski/ Peter Dolanjski] | |||
* Description: How do we decide on the direction to take Firefox in, anyways? Hear about the Core Browser Product Management team's approach to defining Firefox product strategy. | |||
===Products 9: Intro to Statistics via Simulation=== | |||
* Presenter: [https://mozillians.org/en-US/u/rharter/ Ryan Harter] | |||
* Description: Many stats introductions rely on a few magic formulas to get started. In this presentation, we'll lean on some basic familiarity with coding to side step the magic formulas and give a more intuitive introduction to statistics. The talk will likely be around 30 minutes and will include code samples. | |||
===Products 10: Data @ Mozilla=== | |||
* Presenter: [https://mozillians.org/en-US/u/mhoye/ Mike Hoye] | |||
* Description: A series of lightning talks about available data sources, repositories and resources available at Mozilla. | |||
===Products 11: Firefox Growth: The How-To of Gaining Browser Market Share=== | |||
* Presenters: [https://mozillians.org/en-US/u/adavis/ Alex Davis] and [https://mozillians.org/en-US/u/cmore/ Chris More] | |||
* Description: How will we achieve our growth goals for Firefox in 2017? If you are involved with attracting or acquiring new users, building or managing product features, or developing strategies for retaining quality users, please join us to learn about how growth methodologies can positively impact everything from marketing to product. You will learn about the different approaches to growing Firefox's user base, the process of hypothesis-driven experimentation, and practical examples of how we all can contribute to meeting next year's goals. ''Note: no spreadsheets will be harmed in this presentation.'' | |||
===Products 12: Add-on Recommendations Without Add-On Data: Crash Course in Pairwise Similarity Modelling=== | |||
* Presenter: Martin Lopatka | |||
* Description: | |||
<BLOCKQUOTE>Given a new user, with no history of using add-ons in Firefox, collaborative filtering models have no information on which to base recommendations. Alternatively, we can recommend the (globally) most popular add-ons. This feels like an unsatisfactory strategy given the vast information contained in the telemetry corpus that may indirectly help with making personalized recommendations. | |||
In this workshop we will explore two main lines of inquiry: | |||
1. Given the substantial amount of information contained in a telemetry ping, can we leverage other features of new users to help define clever addon recommendations based on analogies to other users. | |||
2. Can this process be used to answer exploratory questions about the relationships between features contained in the telemetry data as well as relationships between firefox users in a variety of other contexts? | |||
We will explore the representation of telemetry data as a set of similarities between user pings relating to their usage of add-ons. Then we will explore other characteristics can be used to predict similarity in terms of add-on usage. Determining a set of surrogate measurements will allow us to make personalized add=on recommendations for users who have never explored the wide world of Firefox add-ons.</BLOCKQUOTE> | |||
=Technology= | =Technology= | ||