The Future of Data Science

Data Science

This year's annual conference in New York of The Association for Computing Machinery, a professional association in computer science, focused on data science, reports The New York Times.

The big data and data mining fields are booming. In the fast-moving field, everyone is eager to keep abreast of the current trends and also look into the future.

In one keynote speech, Oren Etzioni, the chief executive of the Allen Institute for Artificial Intelligence—which is financed by Microsoft co-founder Paul Allen—discussed the limits of the big data approach.

In a presentation titled "The Battle for the Future of Data Mining," he told the audience not to be overly influenced by the big data tidal wave with its emphasis on mining large data sets for correlations, inferences and predictions.

"The big data approach is brimming with short-term commercial opportunity, but scientists should set their sights further." Google and Bing are search engines that provide websites. Wolfram Alpha and IBM's Watson are computational knowledge engines that can provide an answer to a question. But they don't have common sense knowledge, things that are easily understood by humans.

He showed the sentence: "The large ball crashed right through the table because it was made of Styrofoam." He asked, What was made of Styrofoam, the large ball or the table? The table, humans will answer. But the question is a conundrum for a software program, explained Etzioni, because the correct answer involves both grammar and background knowledge. The latter is something humans acquire through experience of the world, which computers currently don't.

Etzioni is skeptical of the progress that is possible with "deep learning," an artificial intelligence technique that uses the structure of the human brain as a model for computer systems that process huge amounts of data. Instead, at the Allen Institute, Etzioni is leading a growing team of 30 researchers that are working on systems that move from data to knowledge to theories which should then be able to reason.  

Another keynote speaker, Eric Horvitz, a computer scientist at Microsoft Research, agreed with Etzioni that the long-range goal is for computer systems that can reason rather than merely recognize patterns and correlations and make predictions.

But in Horvitz's presentation "Data, Predictions and Decisions in Support of People and Society," he chose a different emphasis. "I think we can have a huge impact in so many fields in the shorter term, along the way to deeper machine intelligences."