Difference between revisions of "Earth Science Data Analytics/2016-1-21 Telecon"

From Earth Science Information Partners (ESIP)
 
Line 3: Line 3:
 
===Known Attendees:===
 
===Known Attendees:===
  
ESIP Host (Erin Robinson), Steve Kempler, Tiffany Mathews, Lindsay Barberie, Chung-Lin Shie, Robert Downs, Beth Huffer, Joan Aron, Brian Wee(?), Byron Peters, Abby Benson, H. Joe Lee
+
ESIP Host (Erin Robinson), Steve Kempler, Tiffany Mathews, Lindsay Barberie, Chung-Lin Shie, Robert Downs, Beth Huffer, Joan Aron, Brian Johnson, Byron Peters, Abby Benson, H. Joe Lee
  
  
Line 32: Line 32:
  
  
Excellent discussion…  
+
Excellent discussion… Thanks Bar, for taking great notes.
  
  
Line 39: Line 39:
 
The discussion transformed into what we most recently learned and how we should go forward with it.  That being, of the three types of ESDA, Data Preparation, Data Reduction, and Data Analysis, the latter is the most difficult to develop an approach for, because science research/analysis is very individual; libraries of mathematical tools already exist, and; the plethora of specific research models are unique, and well understand by the researcher, rendering us no real opportunity to add value for large groups of users per model.
 
The discussion transformed into what we most recently learned and how we should go forward with it.  That being, of the three types of ESDA, Data Preparation, Data Reduction, and Data Analysis, the latter is the most difficult to develop an approach for, because science research/analysis is very individual; libraries of mathematical tools already exist, and; the plethora of specific research models are unique, and well understand by the researcher, rendering us no real opportunity to add value for large groups of users per model.
  
The group decided that heterogeneous Data Preparation is where the most pain points are, and tools/techniques that target heterogeneous data preparation should be targeted first.
+
The group decided that heterogeneous Data Preparation is where the most pain points are, and tools/techniques that target heterogeneous data preparation should be targeted first.  Addressing Data Preparation needs will directly help Data Analysis, even if not specific research analysis.
  
With this approach, it was noted that we need to invite more scientists to our cluster to provide more insights to their experiences and needs regarding the co-analysis of heterogeneous data.  Perhaps we should institute a 'science advisory board'.  Like wise, we should include applications researchers, who naturally work with various datasets, often not even in the same discipline, to derive results form their studies.
+
With this approach, it was noted that we need to invite more scientists to our cluster to provide more insights to their experiences and needs regarding the co-analysis of heterogeneous data.  Perhaps we should institute a 'science advisory board' (maybe ESIP would/will)Also, we should include applications researchers, who naturally work with various datasets, often not even in the same discipline, to derive results from their studies.
  
 +
Brian suggested we can invite people from the SMAP science team and ISAT2 'early adapters', two communities Brian is familiar with.  Likewise, Chung-Lin suggested GPM data users.  Airborne data users who also work with satellite data are also excellent candidates,  See action below.
 +
 +
Also, discussed:  Reference to DARWIN-CORE (standards) and ESRI (GIS) tools, both of which facilitate a better understanding of multi-data analysis.
  
  
Line 78: Line 81:
 
Agenda:   
 
Agenda:   
  
Let's invite the ESIP community and present the work of our previous 19 telecons and 5 face-to-faces
+
'''Presentation:  Let's invite the ESIP community and present the work of our previous 19 telecons and 5 face-to-faces'''
 +
 
  
 
===Actions:===
 
===Actions:===
  
1. Steve, Joan, Ethan, Sean, Chung-Lin, Rob, Thomas:
+
1. Brian and Chung-Lin - Please identify one or two multi-data researchers who would be willing to provide insights into their experiences and needs for accessing and preparing data for co-analysis of heterogeneous dataLet's invite them to our March telecom for an informal discussion(If we can get 3 to 5 people, we can have a 'panel' Q&A session)
 
 
a.  Read paper provided by Ethan:  http://www.boozallen.com/insights/2015/12/data-science-field-guide-second-edition
 
 
 
bDescribe the ESDA tools/techniques we identified on our matrix shown above (More details to follow in e-mail)
 
 
 
cMap techniques defined in the boozallen paper to our ESDA goals requirements, as appropriate (More details to follow in e-mail)
 
 
 
  
2.  All other ESDA members: Help… please let Steve know if you can help us with Action #1
+
2.  Steve, others to be contacted: Complete categorization of all identified ESDA tools/techniques by Data Preparation, Data Reduction, and Data Analysis.  (https://docs.google.com/spreadsheets/d/1zMczlWZnQUiubyfcLjwDIQ-SGm7Sm5C2Wo9bc75V-qk/edit#gid=0)

Latest revision as of 19:21, February 6, 2016

ESDA Telecon notes – 1/21/16

Known Attendees:

ESIP Host (Erin Robinson), Steve Kempler, Tiffany Mathews, Lindsay Barberie, Chung-Lin Shie, Robert Downs, Beth Huffer, Joan Aron, Brian Johnson, Byron Peters, Abby Benson, H. Joe Lee


Agenda:

Agenda

1. ESIP Cluster Meeting recap

2. ESDA for Data Preparation, Data Reduction, Data Analysis – Where can Information Technology make the biggest impact?

3. Tools and Techniques – How can we best organize the plethora of tools and techniques we have uncovered?

4. Open Mic – What else should we be addressing?


Presentations:

None, this time.

Use Case Information: https://docs.google.com/document/d/1U1mAt4ZjJqXeNmtRoE4VbI1nBgS1v7DzeHib_7mzOF8/edit


Notes:

Thank you all for attending and participating in our telecon.


Excellent discussion… Thanks Bar, for taking great notes.


Steve started by summarizing the January ESIP ESDA Cluster Meeting. The ESDA Cluster continues to attract people who are interested in learning about data analytics, what it means , and how it 'fits' in the work we do. This has made the cluster, unlike most other ESIP groups, more academic in nature. Data Analytics takes on new methodologies of data analysis born out of the innovative opportunities created by researchers, afforded by the readily available explosion of heterogeneous information. This is new ground, and we are all learning the landscape, ultimately to understand the technology/methodology gaps we can fill. At the Cluster meeting, we discussed our work in collecting use cases, potential Earth science data analytics tools and techniques employed, our AGU study which entails studying the analysis methodologies of several dozen science projects, use cases, and next steps.

The discussion transformed into what we most recently learned and how we should go forward with it. That being, of the three types of ESDA, Data Preparation, Data Reduction, and Data Analysis, the latter is the most difficult to develop an approach for, because science research/analysis is very individual; libraries of mathematical tools already exist, and; the plethora of specific research models are unique, and well understand by the researcher, rendering us no real opportunity to add value for large groups of users per model.

The group decided that heterogeneous Data Preparation is where the most pain points are, and tools/techniques that target heterogeneous data preparation should be targeted first. Addressing Data Preparation needs will directly help Data Analysis, even if not specific research analysis.

With this approach, it was noted that we need to invite more scientists to our cluster to provide more insights to their experiences and needs regarding the co-analysis of heterogeneous data. Perhaps we should institute a 'science advisory board' (maybe ESIP would/will). Also, we should include applications researchers, who naturally work with various datasets, often not even in the same discipline, to derive results from their studies.

Brian suggested we can invite people from the SMAP science team and ISAT2 'early adapters', two communities Brian is familiar with. Likewise, Chung-Lin suggested GPM data users. Airborne data users who also work with satellite data are also excellent candidates, See action below.

Also, discussed: Reference to DARWIN-CORE (standards) and ESRI (GIS) tools, both of which facilitate a better understanding of multi-data analysis.


To Do List:

Done:

1. Finalize ESDA Definition and Goal categories

2. Write letter to ESIP Executive Committee proposing that the ESDA Definitions and Goal categories be ESIP approved

3. Characterize use cases by Goal categories and other analytics driving considerations

4. Derive requirements from #3

Underway:

5. Further validate requirements with (many) more additional use cases

6. Survey existing data analytics tools/techniques

7. Write our paper describing ... all the above


Questions to think about:

What is the best way to record use cases, and associated requirements, and matching tools? A forum?


Next Telecon:

February 18, 2016 ESDA Telecon XX (our 20th telecon)


Agenda:

Presentation: Let's invite the ESIP community and present the work of our previous 19 telecons and 5 face-to-faces


Actions:

1. Brian and Chung-Lin - Please identify one or two multi-data researchers who would be willing to provide insights into their experiences and needs for accessing and preparing data for co-analysis of heterogeneous data. Let's invite them to our March telecom for an informal discussion. (If we can get 3 to 5 people, we can have a 'panel' Q&A session)

2. Steve, others to be contacted: Complete categorization of all identified ESDA tools/techniques by Data Preparation, Data Reduction, and Data Analysis. (https://docs.google.com/spreadsheets/d/1zMczlWZnQUiubyfcLjwDIQ-SGm7Sm5C2Wo9bc75V-qk/edit#gid=0)