Here on Data-Smart, Stephen Goldsmith wrote about the use of chatbots to make 311 call centers more efficient. The North Carolina Innovation Center has begun using chatbots to handle simpler tasks for its internal IT help desk hotline, which frees up IT center workers to focus on more time-intensive service requests. Similarly, 311 centers could use bots to address simple requests, read applicable documents, mine social media, and turn more difficult problem-solving over to engaged operators.

Also on Data-Smart, we highlighted the data-related initiatives that were recognized as Semifinalists in this year’s Innovations in American Government Awards program, which acknowledges novel, effective, and replicable initiatives. Semifinalists included 18F, a digital consultancy inside the U.S. government that works with federal agencies to deploy digital tools; Computer Vision for Conservation, a crowdsourced algorithm used to identify and protect North Atlantic Whales; and the State of Washington’s Regulatory Roadmap Initiative, an easy-to-understand tool outlining regulatory requirements for businesses. We also profiled those data-driven programs recognized as Bright Ideas, which acknowledges noteworthy submissions to the Innovation in American Government Awards program that did not advance to the final rounds of the competition.

The Harvard Kennedy School Government Performance Lab (GPL) released a set of outcome and process metrics recommendations for addressing homelessness. GPL developed these metrics during its What Works Cities work with Seattle’s Human Services Department and offers them as a resource for other jurisdictions seeking to adopt a data-driven approach to managing homeless services contracts. Using these metrics will allow governments to quickly spot problems and opportunities for improvement and operationalize changes to drive better performance during the course of a contract.

GovLab wrote about Manila’s data mining efforts to understand traffic congestion. Three ridesharing companies—Easy Taxi, Grab, and Le.Taxi—are working with the World Bank to make traffic data from their drivers’ GPS streams available to the public through an open data license. This data will help resource-constrained transport agencies make better, evidence-based decisions about traffic signal timing plans, public transit provision, and roadway infrastructure needs, among other things.

Forbes discussed a partnership between the Blue Cross Blue Shield of Tennessee (BCBST) and data analytics company Fuzzy Logix to tackle the opioid crisis using data. The project leveraged years’ worth of pharmacy and claims data to identify people at risk of developing an opioid abuse problem. Analysts determined that clusters of behaviors like frequent use of different prescribers and dispensaries could predict later abuse or misuse issues with up to 85% accuracy.

On the Sunlight Foundation blog, Howard Lim, project manager of Boston’s Open Data to Open Knowledge initiative, detailed the city’s efforts to host conversations on how to share open data with citizens in a meaningful way. Takeaways from these conversations included the need to sharpen communications about what open data is and what citizens can do with data, as well as the revelation that libraries, as a pillar of community life, can increase public interest in open data.

The GovLab at NYU Tandon School of Engineering has launched a new website, DataCollaboratives.org, to highlight public-private partnerships in which the participants exchange data to solve public issues. To take full advantage of the recent growth in data collection, it is critical that governments and private companies alike share their information with one another. Data collaboratives can provide better situational awareness and response, improved public service delivery, enhanced knowledge creation and transfer, cutting-edge prediction, and more effective impact assessment.