Airbnb open sourcing Airflow, Aerosolve for machine learning, data discoveries

Airbnb also debuted another pair of new features catering more to front end users, meaning hosts and guests.


Airbnb is going open house on open source with a pair of new projects that double down on all that traveling data moving in.

Machine learning face-off: Microsoft uses Band to show what its Watson rival is capable of

Plug-and-play machine learning data models that can instantly analyze information and provide intelligent insights? Yep, Microsoft has 'em, it says.

Read More

Unveiled at the company's OpenAir engineering summit in San Francisco on Thursday, the vacation rental wunderkind announced it was open sourcing two of its homegrown data mining products: Airflow and Aerosolve.

Airflow is a workflow management platform built for authoring, scheduling and monitor data pipelines at scale in a timely manner -- an absolute necessity for a burgeoning global travel service that has exploded in the last few years.

According to Airbnb, Airflow has helped streamline many processes for its internal data engineering team.

But the bigger opportunity for outsiders looking into Airbnb's secret lair is Aerosolve, a machine learning package for big data engine Apache Spark designed to improve how people interpret complex data sets.

For instance, Airbnb's engineering team boasted in a blog post that Aerosolve has enabled Airbnb to identify and better understand the relationship between the price of an Airbnb listing in a given market and factor in demand for that listing.

Airbnb engineers explained further:

Many features go into predicting the demand for a listing among them seasonality, unique features of a listing and price. These features interact in complex ways and can result in machine learning models that are difficult to interpret. So we went about building a package to produce machine learning models that facilitate interpretation and understanding. This is useful for us, developers, and also for our users; the interpretations map to explanations we provide to our hosts on why the demand they face may be higher or lower than they expect.

Aerosolve is available via Github now.

Airbnb also debuted another pair of new features catering more to front end users, meaning hosts and guests.

Those would be online gift cards (pictured above and available only in the United States initially) and a tool designed to help hosts with configuring pricing.

Dubbed Price Tips, the recommendation feature provides Airbnb hosts with ongoing advice on how to price listings for each day of the year with the goal of bumping up booking conversion rates.

Price Tips spits out advice based on an assortment of factors, including average demand, location, home/space rental type and other travel trends.

Not by coincidence, Price Tips is powered by Aerosolve, reflecting a rapidly growing interest and investment in machine learning not just by Airbnb but many tech companies these days from Pinterest to Amazon Web Services.

Airbnb seems like a business just ripe for machine learning harvesting with a multitude of different metrics streaming in -- all through a platform based in the cloud with many of its users accessing it via mobile devices.

With the opportunity for quantifiable trends and insights in real-time and being able to offer that nearly immediately to its users, machine learning could provide quite the edge for Airbnb in the travel industry at large.

Image via Airbnb