Merge pull request #13 from marda-alliance/PeterKraus-patch-1

Minutes from progress meeting (2023-01-19)

meetings/2-meeting-2023-01-19/minutes.md

@@ -0,0 +1,71 @@
+# Progress meeting
+
+19/01/2023
+
+## Agenda
+
+- Introductions (5 mins)
+- Summary of main goals (Matthew Evans)
+- Progress in WP1, WP2, WP3 (Matthew Evans)
+- Case study: yadg (Peter Kraus)
+- Open discussion
+
+## Minutes from Discussion
+
+1. David Elbert:
+  - Suggestion: Please register for the [annual MaRDA meeting](https://www.marda-alliance.org/blog-2/marda2023/). Poster sessions for ECRs are scheduled, with prizes!
+
+---
+
+2. Matthew Evans & Peter Kraus: 
+  - Presentation of current progress and case study on [yadg](https://github.com/dgbowl/yadg) ([slides](https://docs.google.com/presentation/d/1nQD7cwEG67W5MAVjXXb-6nUklDg1AcxYs8j7WRsJjIc/edit#slide=id.g1e03a09dbfb_0_16))
+
+2. a) Casper Andresen:
+  - Question: Will the transformation step in our pipeline be based on semantics, or only a pure syntax translation?
+
+2. b) Matthew Evans:
+  - Answer: LinkML allows linking to other schemas, such as Dublin Core or Schema.org. Therefore, capturing semantics should be possible. Contributions from experts are welcome!
+
+---
+
+3. a) Logan Ward:
+  - Question: Interesting choice of going with a "dataschema" for yadg. What exactly is it? What are the limitations?
+
+3. b) Peter Kraus:
+  - Answer: It defines the "folder structure" of files related to a single experiment, as opposed to the layout within the individual files. User adoption is quite hard, and very often you just want the raw data, without bundling it.
+
+3. c) Logan Ward:
+  - Suggestion: Efforts such as this one (metadata extractors) could lead to a nice recursion, where your extractor taps into other linked tools.
+
+---
+
+4. a) Steffen Brinckmann:
+  - Question: Is passing parameters into extractors (e.g. information about background etc.) condsidered for the schemas?
+
+4. b) Peter Kraus:
+  - Answer: From my point of view, not at the extractor level, which should not be post-processing the data values.
+
+4. c) Matthew Evans:
+  - Answer: This is a very application-dependent question. Parameters are likely to strongly depend on the extractors themselves. If we support it, it must be optional.
+
+---
+
+5. a) Ken Kroenlein:
+  - Suggestion: For extractor output, the data dictionaries WG of MaRDA can provide a controlled vocabulary.
+
+5. b) Peter Kraus:
+  - Answer: I will be in touch to figure out how I can implement this in my code!
+
+---
+
+6. Closing remarks:
+- MaRDA annual meeting on 25. - 27. Feb. 2023
+    - have a first tagged schema by then (ME)
+    - have an example implementation in yadg (PK)
+- Office hours: 
+    - fortnightly (approximately), informal
+    - next one on 25. Jan 2023, 15:00 UTC
+    - suggest a date/time if the dates planned don't suit!
+
+
+