Big Data, Big Impacts in Sustainability Webinar Questions

What software is being used for analysing data? Vensim? Powersim? Sysdea? Are any of these tools open source? if so, can you give us a resource list?

We use a lot of open source software (such as R) as well as commercial software (such as Matlab or Tableau).

Open Source Software for Data Science:

• R & R-Shiny
• Google Charts
• Python: numpy & scipy

GIS – Mapping software

• https://www.arcgis.com/home/index.html (free for a few months)
• https://www.google.com/earth/download/gep/agree.html

Commercial Software (Free trials available)

• https://www.mathworks.com/products/matlab.html
• https://www.tableau.com/

We create our own programs. Python is an extensively used high-level programming language for general purpose. Its design highlights code readability and its syntax allows analyst to express concepts in fewer lines of code that would be possible in languages such as C++ or Java.

Can you please elaborate on the riskiness of using FB for political campaigning?

This FB alleged political campaigning and advertising is now in front of congress. We have to wait and see how social media, which is now the main source of news for many people, will be regulated by FCC.

Social Media, and specifically Facebook has been so successful because it provides a “feed” of content that is interesting and relevant to you. However, part of the dangers of social media is that you build your own echo-chamber of interests. This has contributed to the hyper-polarization we have seen in politics in general. There is no good way to fix this aside from being aware of the problem and seeking out diverse ideas and opinions.

Could we possible use trees to spot gas pipe’s leaks?

• We are working with expert arborists who can tell you if the tree is dying. But experts validate the arborist data with actual measurement of methane gas using many types of devices including Picarro gas analyzer. Ultimately, we can get enough tree samples and train a neural network to predict. Note we need really good training data.

Check Picarro gas analyzer here.

Could you speak to what 2-3 undergraduate courses (e.g., Intro GIS, Programming, etc) would be most valuable to develop/teach for UGs to participate in Big Data careers?

You need data programming, data visualization and machine learning types of classes. Here are some examples

• • R programming or AI or data visualization course:
  • https://www.coursera.org/specializations/data-visualization
  • https://www.coursera.org/learn/r-programming
  • https://www.coursera.org/learn/machine-learning
  • GIS Programming classes – Check your local schools – several offer the courses. You can also take it online from the company at: http://learn.arcgis.com/en/projects/get-started-with-arcgis-online/

Are you aware of any big data projects tracking ecosystem threshold data?

We are not sure what you are asking. NSF has funded cyber enabled projects for various ecosystems including oceans and coastal areas. https://www.nsf.gov/funding/pgm_summ.jsp?pims_id=504829

How can you estimate the statistical power of the predictions in these models eg in spatially explicit ecosystem services models?

We use expert validation or Monte Carlo Simulation which provides an estimate of the probability distribution of possible values that is derived from uncertainty about the underlying parameters and relationships. We can test our results against this distribution given our data distribution may not be a normal distribution. There is ongoing work in this area. Stay tuned!

How did you test the model in the Malaria case?

Testing or validation requires out of sample testing (Unseen test data) or validating with experts who have knowledge or understanding of what is happening on the ground.

The research presented here, particularly the case studies, obviously have great practical applications. Are they all done by Chatmine Technologies as projects for specific clientele? If not, how receptive are stakeholders and government organizations to the results of your studies (i.e. ones primarily done by Boston U?)

Our Malaria project in Ethiopia and food security in Cambodia are both done as part of BU projects. Social media analysis and municipal dashboard products are Chatmine projects. Visit Suchi’ academic page here

There are something over 10,000 new peer-reviewed published studies in Climatology every year; Hayhoe et al (2016) and others do analyses of samples from these papers; are any of the tools discussed appropriate to improving survey of papers?

We were not able to find the paper you are referring to. In general, meta analysis of peer reviewed articles are very important in medicine and every other science. It tells us what can be concluded when we combine many studies together done under varying contexts. I have done two such papers and learned that in science we may ignore regions of the world due to lack of funds or learn something new (e.g. what has not investigated yet) and novel worth exploring. All PhD Students are encouraged to read in their areas to get a more detailed understanding of consensus in their fields as well as areas worth exploring.

What about review human subjects protection when working with de-identified big-data files? What if you are not affiliated with institution with review board?

We suggest that it would be wise to partner with a local university to do this research. You cannot do research involving human subjects without Institutional Review Board (IRB) approval. No journal would publish the results.

What ethical guidelines for privacy and other ethical concerns do you follow?

Most of the work presented here involving human subjects was done at Boston University with prior IRB approval; Chatmine business data has very strict guidelines for sharing and we don’t share any of our client project data without prior approval from their legal teams. We strictly follow privacy standards and do not collect or keep any personally Identifiable Information (PII). We absolutely adhere to ethical and legal standards in any work we do.

How do you think the trend towards sustainability is impacting individual experiences and lifestyle? Are you harmonizing or using any of the established ISO metrics for urban resilience and sustainability? I joined late but wondered if you mentioned the reception and adoption of these tools by municipal decision-makers and outcomes.

• Suchi teaches a class on Sustainable Supply Chains at the Vienna University of Economics and Business Administration. ISO standards are followed widely in the European Union. In the US, there are other standards such as Fair Trade (Starbucks). We can design our metrics to fit each standard. That’s a great suggestion since it opens up possibilities. We will contact you.
• We are developing this product for country of Cambodia and will present our results to stakeholders including the ministry next year. We have not used formal standards yet. Suchi has developed an example around sustainable fish paste production that Cambodian women engage in. Again standards are good to use to frame the work.

Are there municipal dashboards in operation in any cities in California or at the State level?–Where can we go to see more demonstrations of these “Urban pulse” dashboards?

This is our proprietary Chatmine Product that we can make for any other ste using open data from various sources. So please contact us to set a time to talk if you have interest in this product.

Would the municipal dashboard example scale to other states?

This is our proprietary Chatmine Product that we can make for any state using open data from various sources. So please contact us to set a time to talk if you have interest in this product.