Wolfram Alpha Makes Alexa Smarter
Last week, Alexa announced it’s integration with Wolfram Alpha, a computational knowledge engine that offers a wide range of curated data, focusing on math and science. Many people think of Wolfram Alpha as the leading search engine for structured data sources such as statistics about a variety of topics ranging from population and economics to sports and medicine. With its robust search capabilities, Google Assistant has historically held the title of best performing smart speaker when it comes to answering questions. Alexa has been working to close that gap by recently integrating Wikipedia, Yext, Accuweather, and Stats.com. To test out the Wolfram Alpha integration ask, “Alexa, how fast is the wind blowing right now?” or “Alexa, what is the billionth prime number?” An Amazon spokesperson told VentureBeat,
We rolled out an Alexa Q&A integration with Wolfram Alpha to U.S. customers, which expands Alexa’s capabilities to answer more questions related to mathematics, science, astronomy, engineering, geography, history, and more. Information curated by Wolfram Alpha has rolled out to select customers and will continue to roll out over the coming weeks.
What Wolfram Alpha Brings to Alexa
Wolfram Alpha was launched in 2009 as a computational knowledge engine, also known as an answer engine. Rather than providing a list of documents or web pages that might contain an answer, as search engines do, the answer engine provides solutions to factual queries by actually computing or presenting the answer. Externally sourced “curated data” is used to compute answers. Data is collected from the CIA’s The World Factbook, the United States Geological Survey, Dow Jones, and Crunchbase among other sources.
The curated datasets are checked for quality either by a scientist or other expert in a relevant field. Responses can include relevant visualizations and live datasets, too. For example, Wolfram Alpha can use the live dataset of a consenting Facebook user’s profile to generate a “personal analytics” report. This report contains the age distribution of friends, the frequency of words used in status updates, in addition to other information. Overall, Wolfram Alpha brings a lot of new capabilities and knowledge access to Alexa.
Alexa Wants to Maintain Market Share
Alexa’s strategy to be ‘everywhere’ has definitely worked out in terms of gaining the smart speaker market share in the U.S. and U.K. Voicebot.ai found in September 2018 that Amazon’s installed base market share for smart speakers was 64.6%. Amazon is looking to maintain that lead and has been extending Alexa’s capability to close the gap in areas where it trails its peers.
Alexa has partnered with numerous third-party companies to compensate for one of the biggest aspects it has been shown to lack: a strong knowledge base. A Loup Ventures study from December showed Google Assistant was able to answer 88 percent of 800 questions correctly. In comparison, Siri answered 75 percent correctly, and Alexa answered 73 percent correctly. Google Assistant has a unique advantage in that the Google search engine provides a wealth of knowledge for the assistant. This allows Google Assistant to consistently dominate other voice assistants when comparing user query performance – the most popular use case for smart speakers.
Third-Party Partnerships Help Alexa Close the Knowledge Gap
In an effort to improve its quality, Alexa announced the integration of Yext just this summer. Yext provides access to the hours of operation, descriptions, and addresses of partnered businesses. Answer Updates rolled out in August, providing users with a notification when Alexa learns the answer to a question that is previously could not answer. Alexa Answers was then launched in early December, inviting users to provide responses to still unanswered user queries. Bill Barton, Vice President of Amazon’s Alexa division, said earlier this month,
Our vision has always been that Alexa will be able to answer all questions in all forms, from anywhere in the world.
Although still trailing other voice assistants in the Loup Ventures study from this month, it is important to note that in comparison to a Loup Ventures study conducted in July, Alexa has actually improved by 11.6 percent. A significant factor of this improvement can be attributed to Alexa’s third-party partnerships that provide datasets Alexa would otherwise not have access to. It will be interesting to see how the integration of Wolfram Alpha will impact Alexa’s query performance, especially considering that Siri integrated Wolfram Alpha back in 2011.