Thesis Meeting | Sep 17, 2024 - 2:30pm - *ITB 167*

[samm82]

Agenda

In-progress Work

May not be discussed in meeting if it's better to do asynchronously/if there's nothing to discuss yet.

1. Methods section: https://github.com/samm82/TestGen-Thesis/pull/91
2. Sources section: sources branch
3. Ideal workflow
   1. Should I open one PR at a time, or can I open multiple? Can they be based off of each other?
   2. Many sections are interrelated; how should I request reviews of sections that are updated in tandem?
   3. How to track future work in a section; is it easier to re-review the whole section or just the changed parts?

Classifying Discrepancies (reclassDiscreps branch)

1. New categories:
   1. Definitions (for incorrect, missing, inconsistent, ambiguous, or overlapping definitions for approaches or related terms)
   2. Terminology (for incorrect usage of or inconsistencies with terms, such as an acronym with two meanings, esp. in the same doc!)
   3. Sources (for issues with sources cited by the document itself; usually claims that a source defines something when it doesn't)
2. Removal of the "other" discrepancies section! 🥳
3. Future work:
   1. Write up explanations for these categories
   2. Restructure discrepancy list based on these changes (i.e., sort by discrepancy type, not severity; remove severity discussion)
   3. Pie chart!
   4. Decompose some of these sections (if not in the table, in the sections themselves?)

Thesis Next Steps

1. How should I write about/discuss my research question(s) in my thesis from a structural standpoint? My NEW ones are:
   1. What testing approaches are described in the literature?
   2. What discrepancies exist between descriptions of these testing approaches?
   3. Can any of these discrepancies be resolved/reduced systematically?

[samm82]

Thesis Writing

• [ ] Look more into thesis-writing guidance
  • [x] "Background" should be "Introduction"
  • [ ] Introduction: what is the problem, and what am I doing about it? (look for how to write an introduction for a paper)
    • [ ] Can include the history of this thesis, explaining how/why it is presented "out of order" (e.g., why definitions arose from analysis but were present before)
    • [ ] Can provide examples of "bad" taxonomies
  • Don't be afraid to ask about the merit of specific resources!
  • [ ] Contributions section: how much did I do?

Workflow Guidance

• [ ] Asking for review should be done on a given PDF (with link) INDEPENDENTLY of PRs
• [ ] Assign a SPECIFIC person for a review, then the other, vs. both at once

Formalization

• [ ] Formalize analysis/counting (discrete math)
  • [ ] We expect a tree, so cycles don't make sense
  • [ ] For definition discrepancies, what are properties of "good" definitions?
    • [ ] Maybe call "discrepancies" something stronger, like "failures" or "violations"