The Power of Data Science and AI for Social Good

Ubiquitous sensors and sources everywhere deliver deep and diverse data as the new fuel for value creation in nearly every industry and organization. The bonanza of digital benefits includes insightful discoveries, smarter decisions, essential innovations, and better outcomes.

Since the early days of the current big data, data science, and artificial intelligence (AI) technology revolutions, many data scientists have been motivated (even compelled) to apply these data collections and their skills with advanced algorithms to solving important problems with significant societal impact.

In some organizations, internal data hackathons are used to encourage use (and quality improvements) of their data collections, to tackle difficult business problems (both for the organization itself and for its clients), and to inspire innovative solutions and new ideas that might inform the development of new business strategies or new products.

“Data for good” and “AI for social good” are rallying cries for numerous data hackathons, open data initiatives, datafests, codefests, and machine learning challenge competitions worldwide. All are welcome (and encouraged) to participate, regardless of skill level. The participants are very willing to trade a few hours on a weekend to learn new techniques, solve challenging problems, explore interesting datasets, develop potentially valuable solutions, and share some pizza with new friends.

Pizza with friends is great, but has limited geographical reach. Conversely, online challenge competitions—which crowdsource hard problems in need of an AI machine learning solution across massive data collections—tend to have the largest participation. The reason for the larger participation is primarily because they are not confined to a specific day or location, are online with global reach, and the challenge problem typically has rich societal consequences. These competitions encourage collaboration, code-sharing, open discussion, and innovation on a phenomenal worldwide scale. The United Nations’ Data for Climate Action Challenge is one example.

Many machine learning challenge competitions offer prizes (including cash for the top finishers), and that helps to draw the crowds. However, participants would come anyway since the positive societal outcomes are a far more lasting and satisfying reward. Furthermore, the innovative solutions and skills enhancement that result from these competitions provide powerful outcomes for the participants as well as for the competition hosts.

Among the major online data science and AI for social good challenge competitions, one of the most successful is the annual Data Science Bowl, presented by Booz Allen Hamilton and Kaggle data science competition platform, and supported by several data and technology partners. Started in late 2014, the Data Science Bowl has addressed challenge problems, involving big data collections and complex machine learning tasks, in ocean health, heart health, early detection of lung cancer, and faster discovery of new drugs to treat common and rare diseases.

One of the significant benefits and consequences that the Data Science Bowl shares with other hackathons and challenge competitions is the diversity in its participants, algorithms developed, and solutions delivered. Innovation is a natural consequence of opening up the datasets and posing hard questions to a community of skilled, motivated, and diverse participants.

Finally, we underscore the power of open data. It is the essential input to data science competitions, rich informant for smart analytics innovations, and necessary fuel to drive AI applications. Open data repositories are data scientists’ dream in so many ways! Open data repositories:

1. Provide a source of insight and transparency into the domains and organizations that are represented by the data sets.
2. Enable value creation across a variety of applications, using the data as the basis for innovation, transformation, new ideas, and new businesses.
3. Offer a rich variety of data sets for data scientists to sharpen their data mining, knowledge discovery, and machine learning modeling skills.
4. Allow for many more eyes to look at the data and thereby to see things that might have been missed by the creators and original users of the data.
5. Enable numerous “data for social good” activities (e.g., hackathons, citizen-focused innovations, public development efforts).

Therefore, when considering how one can impact outcomes for social good and innovate in the rapidly evolving worldwide digital economy, one need look no further than the power of data science and AI to impel, inform, and ignite such opportunities.