Difference between revisions of "2017-2018 Speaker Series"
Annieburgess (talk | contribs) |
|||
Line 1: | Line 1: | ||
+ | '''04/18/2018 Rick Allmendinger''', Professor of Earth & Atmospheric Sciences at Cornell University and several years mapping with the US Geological Survey. Abstract: Abstract: Adventures of a Field Geologist in Academic Software Development: As someone whose professional career coincided with the dawn of the age of desktop computers to the dawn of the age of mobile devices, I've had to write most of my own software for the research that I wanted to do. While developing the algorithms and the software was a necessity, making it pretty and easy to use has been something of a hobby akin to woodworking, which I also enjoy. I discovered early on that, if the software is easy to use and has a nice interface, one is likely to spend more time in the program probing a broader parameter space and asking "what if" questions. All of my programs were developed initially for my own use but some have become quite popular as they have been released for free on my web page. The most popular, Stereonet, is a program for producing stereographic projections that structural geologists love so much. This program has been in nearly continuous development for the last 33 years and is currently available for Macintosh, Windows, and Linux platforms. My web page for downloads gets ~30,000 hits per year and is is used not only by structural geologists, but also geophysicists, mineralogists, sedimentologists, and even archeologists who use it to plot and analyze the alignment of artifacts. I have several other major programs that have thousands of downloads per year as well as several popular utility programs to enable things like latitude-longitude to UTM conversions. In the last year, I have been exploring the versatility of sensors on smart phones and tablets for collecting data in the field and have now have three apps published in the iOS App Store. Both desktop and mobile versions of Stereonet can act as a lightweight upload client for the StraboSpot database. The field geologist can use the device orientation on the outcrop to measure the location and attitude in space of geologic features and then upload their data to StraboSpot whenever they have a cellular data connection. This talk will focus on my development process, lessons learned (i.e., mistakes made) and thoughts and concerns about the future of academic software development. [XXX Meeting Minutes] [XXX VIDEO] | ||
+ | |||
'''03/21/2018 Yolanda Gil''', Director of Knowledge Technologies and Associate Division Director at the Information Sciences Institute of the University of Southern California, and Research Professor in Computer Science and in Spatial Sciences. Abstract: The goal of the NSF-funded EarthCube OntoSoft project is to create a germinal ecosystem for software stewardship in geosciences that will empower scientists to manage their software as valuable scientific assets in an open transparent mode that enables broader access to that software by other scientists, students, and decision makers. We developed the OntoSoft ontology for describing scientific software metadata, which is being used in APIs and mapped to popular software repositories such as GitHub and Zenodo. Based on this ontology, we designed the OntoSoft distributed scientific software registry that contains more than 600 entries operated by several science communities, and allows searching and comparing software entries across metadata fields. We have also created the OntoSoft Geoscience Papers of the Future initiative to disseminate best practices of software scholarship and reproducible research. Looking towards the future, we are interested in creating semantic descriptions of geosciences models that reduce the time to find, understand, compare, and reuse models. To that end, we are currently extending OntoSoft to create rich model catalogs that include machine-readable semantic metadata for: 1) exposing model variables and their relationships; 2) capturing the semantic structure of model invocation signatures based on functional inputs and outputs and their correspondence to model variables; 3) associating models with readily reusable workflow fragments for data preparation, model calibration, and visualization of results. [https://docs.google.com/document/d/1Von1SXZV9eT1cNokgIzTYGYAZP1qvT7wzqcazjPA2Ss/edit?usp=sharing Meeting Minutes] [https://youtu.be/9LSfxFAI8zI VIDEO] | '''03/21/2018 Yolanda Gil''', Director of Knowledge Technologies and Associate Division Director at the Information Sciences Institute of the University of Southern California, and Research Professor in Computer Science and in Spatial Sciences. Abstract: The goal of the NSF-funded EarthCube OntoSoft project is to create a germinal ecosystem for software stewardship in geosciences that will empower scientists to manage their software as valuable scientific assets in an open transparent mode that enables broader access to that software by other scientists, students, and decision makers. We developed the OntoSoft ontology for describing scientific software metadata, which is being used in APIs and mapped to popular software repositories such as GitHub and Zenodo. Based on this ontology, we designed the OntoSoft distributed scientific software registry that contains more than 600 entries operated by several science communities, and allows searching and comparing software entries across metadata fields. We have also created the OntoSoft Geoscience Papers of the Future initiative to disseminate best practices of software scholarship and reproducible research. Looking towards the future, we are interested in creating semantic descriptions of geosciences models that reduce the time to find, understand, compare, and reuse models. To that end, we are currently extending OntoSoft to create rich model catalogs that include machine-readable semantic metadata for: 1) exposing model variables and their relationships; 2) capturing the semantic structure of model invocation signatures based on functional inputs and outputs and their correspondence to model variables; 3) associating models with readily reusable workflow fragments for data preparation, model calibration, and visualization of results. [https://docs.google.com/document/d/1Von1SXZV9eT1cNokgIzTYGYAZP1qvT7wzqcazjPA2Ss/edit?usp=sharing Meeting Minutes] [https://youtu.be/9LSfxFAI8zI VIDEO] |
Revision as of 08:44, April 18, 2018
04/18/2018 Rick Allmendinger, Professor of Earth & Atmospheric Sciences at Cornell University and several years mapping with the US Geological Survey. Abstract: Abstract: Adventures of a Field Geologist in Academic Software Development: As someone whose professional career coincided with the dawn of the age of desktop computers to the dawn of the age of mobile devices, I've had to write most of my own software for the research that I wanted to do. While developing the algorithms and the software was a necessity, making it pretty and easy to use has been something of a hobby akin to woodworking, which I also enjoy. I discovered early on that, if the software is easy to use and has a nice interface, one is likely to spend more time in the program probing a broader parameter space and asking "what if" questions. All of my programs were developed initially for my own use but some have become quite popular as they have been released for free on my web page. The most popular, Stereonet, is a program for producing stereographic projections that structural geologists love so much. This program has been in nearly continuous development for the last 33 years and is currently available for Macintosh, Windows, and Linux platforms. My web page for downloads gets ~30,000 hits per year and is is used not only by structural geologists, but also geophysicists, mineralogists, sedimentologists, and even archeologists who use it to plot and analyze the alignment of artifacts. I have several other major programs that have thousands of downloads per year as well as several popular utility programs to enable things like latitude-longitude to UTM conversions. In the last year, I have been exploring the versatility of sensors on smart phones and tablets for collecting data in the field and have now have three apps published in the iOS App Store. Both desktop and mobile versions of Stereonet can act as a lightweight upload client for the StraboSpot database. The field geologist can use the device orientation on the outcrop to measure the location and attitude in space of geologic features and then upload their data to StraboSpot whenever they have a cellular data connection. This talk will focus on my development process, lessons learned (i.e., mistakes made) and thoughts and concerns about the future of academic software development. [XXX Meeting Minutes] [XXX VIDEO]
03/21/2018 Yolanda Gil, Director of Knowledge Technologies and Associate Division Director at the Information Sciences Institute of the University of Southern California, and Research Professor in Computer Science and in Spatial Sciences. Abstract: The goal of the NSF-funded EarthCube OntoSoft project is to create a germinal ecosystem for software stewardship in geosciences that will empower scientists to manage their software as valuable scientific assets in an open transparent mode that enables broader access to that software by other scientists, students, and decision makers. We developed the OntoSoft ontology for describing scientific software metadata, which is being used in APIs and mapped to popular software repositories such as GitHub and Zenodo. Based on this ontology, we designed the OntoSoft distributed scientific software registry that contains more than 600 entries operated by several science communities, and allows searching and comparing software entries across metadata fields. We have also created the OntoSoft Geoscience Papers of the Future initiative to disseminate best practices of software scholarship and reproducible research. Looking towards the future, we are interested in creating semantic descriptions of geosciences models that reduce the time to find, understand, compare, and reuse models. To that end, we are currently extending OntoSoft to create rich model catalogs that include machine-readable semantic metadata for: 1) exposing model variables and their relationships; 2) capturing the semantic structure of model invocation signatures based on functional inputs and outputs and their correspondence to model variables; 3) associating models with readily reusable workflow fragments for data preparation, model calibration, and visualization of results. Meeting Minutes VIDEO
02/21/2018 Steve Richard, adjunct research scientist at LDEO (Lamont-Doherty Earth Observatory): Adventures in software development for geoscience data entry and discovery Abstract: As a geologist producing digital geologic maps and databases, and a technical manager for the Arizona Geological Survey (AZGS), National Geothermal Data System (NGDS), and Interdisciplinary Earth Data Alliance (IEDA), I have been involved in projects developing and utilizing software to automate and document geoscience datasets over more than 25 years. This presentation will discuss several specific software projects and their outcomes (to date). These includes a data entry tool for a prototype USGS-AASG National Geologic Map Database project (C++ development), development of tools for streamlining digital geologic map production at the Arizona Geological Survey (ArcGIS extensions), and metadata catalog development using open source software (Geoportal, CKAN). Meeting Minutes VIDEO
01/17/2018 Doug Walker University of Kansas and the StraboSpot Project Abstract: The StraboSpot Project is an initiative by field and laboratory geologists to develop data management and sharing for field data and derived laboratory measurements. The project involves extensive community outreach in early phases to develop methods and interfaces appropriate to and in concert with scientific workflows and domain specific languages (in a geologic sense). We have completed work with structural geologists and are in the early phases of working with petrologists and sedimentologists. The data collection is done on mobile devices, using a hybrid development environment, and on a web application leveraging that code. We use a backend graph database system because of its flexibility to expand to user needs and performance for exploring related observations that span many spatial levels. The presentation is designed to give an appreciation for the system, the domain development strategy, and to invite participation in the project and interactions with the system. Meeting Minutes VIDEO
09/20/2017 Colin Smith of Environmental Data Initiative Abstract: The Environmental Data Initiative will be initializing a code library for information management in the environmental sciences.The library will be a resource to those in need of software solutions to their information management needs and serve as a collaborative space for development of IM software. We intend this project to become a community owned and operated resource so inclusion of community members in the design process is critical. We welcome your feedback to improve the design before it is implemented. VIDEO.
05/15/2017 Kate Keahey is a scientist at Argonne National Laboratory and a Computation Institute fellow at the University of Chicago. Her research interests focus on virtualization, resource management, and cloud computing. She presents Software X: A new open-access software Journal. VIDEO.