Self-driving cars, web search, retail recommendations. Evolving technological advancements in the field of artificial intelligence and machine learning affect our daily lives in many ways. Machine learning is an evolutionary step forward in computer science, with roots in pattern recognition and computational learning theory. In essence, it examines how computers act without being directly programmed. Last year, Amazon launched the Amazon Machine Learning Service, making the personalization efforts it’s known for available to everyday developers.
Amazon’s personalization excellence
Amazon has always been a leader in the online shopping experience, and machine learning plays a big role in its personalization efforts—mainly its ability to recommend groupings of product or upsells based on how it interprets, or learns, your shopping habits. That being said, the recommendation system used by Amazon relies heavily on machine learning to consume big data. Just some of the data points collected by Amazon include previous purchases, product view history, and to an extent geo targeting in terms of shipping call to actions—all of which are used in its machine learning process. The engine itself is then responsible for suggesting recommendations for products consumers may want to buy as they browse the website. Focusing on personalization as a key element proves to be an effective way to increase consumer purchases, and subsequently revenue, for Amazon and all of its online retailers.
Recommendation services for the everyday developer
Amazon’s recommendation system is at the core of their online shopping experience. For the first time, with the Amazon Machine Learning Service, the everyday developer will be able to peek behind the curtains to see what makes this recommendation system so effective and put this to use themselves.
This service is designed for software developers as a simple introduction to machine learning. The service will allow everyday developers who have not specialized and have little to no experience in the complex field of machine learning to use Amazon’s tool to advance their own businesses. The user will be able to use Amazon’s servers and storage to bypass the challenging process of setting up their own infrastructure and easily build online applications to analyze their data.
While this machine learning service is not identical to what Amazon is using itself, the service will still be very valuable to developers. As Matt Wood, a data scientist at Amazon, told Wired magazine, this service is focused on “real-world problems for developers.”
Core business decisions are now made less with gut feelings but more through applications of data science. Giving developers the chance to benefit from machine learning software and create predictive models from their data will undoubtedly put Amazon’s web services at the forefront of data analytics software and benefit business in the long run.
What real users think
A review of the service, written by the FlyDataTeam, says that the “ease at which we could get the prediction model (without thinking too much about the algorithm) was quite impressive” and that the service was able to process large data sets (FlyDataTeam 2015). Another review by Serdar Yegulalp, a senior writer for InfoWorld, discusses the pricing model of this service. Pricing for the service is “based on the time needed to analyze data and to build models, as well as by how many predictions are requested.” However, users will need to pay a premium for real-time pricing.
Despite a few critiques, overall, users seem to be satisfied with this new service and are excited to see what Amazon does next.