# Publications *This chapter is based on the notes [2022-05-10_SPSE_Publication.pdf](../../course_notes/2022-05-10_SPSE_Publication.pdf)* ## Open Access and Open Science - People **always** have been asking for **access** to knowledge - But access is limited through society and technology. Examples for knowledge transfer: - 450 BCE: [Laws of the Twelve Tables](https://en.wikipedia.org/wiki/Twelve_Tables). Written and publicly accessible law - Science: Lectures - Bible: Translation 1522 (to understand), Printing 1452 (to spread) - Technology: Patents and Documentation Aspects that regulate access: - Availability (libraries, internet) - Authorship ("Urheberrecht") - Copyright ("Verwertungsrecht") - Personality rights - Patents Licenses determine the given rights for published works. ### Software: - Closed Source - secret - proprietary - freeware/shareware - Restricted Open Source - Open Source - BSD, MIT, Apache 2.0 - GPL - Careful: open != free ### Papers: * Access Restricted ("pay wall"): subscription or payment necessary * Green Open Access: Authors are allowed to publish their work **on their own homepage** (non-commercial) * Gold Open Access: - Everybody can access the work - Different copyright and licensing models: - Copyright with author **or** publisher - Publisher may hold exclusive rights/license - Publisher may hold non-exclusive license; author can share other licenses on their own Examples for non-exclusive licenses: - [CC0](https://creativecommons.org/share-your-work/public-domain/cc0/) (similar to 'public domain') - [CC BY](https://creativecommons.org/licenses/by/2.0/) standard - [CC BY-SA (shared alike)](https://creativecommons.org/licenses/by-sa/2.0/) - [CC BY-NC (non-commerical)](https://creativecommons.org/licenses/by-nc/2.0/) - [CC BY-ND (no derivatives)](https://creativecommons.org/licenses/by-nd/2.0/) Why should you want to transfer copyright to a publisher? * Publisher demands it; no other option or expensive. * Publishers protect the rights of the paper. ## Publishing Fees Publishing is expensive. So why publish with a publisher, and not yourself? - Peer-reviewing often organized by publisher or conference - Reputation - Copy-Editing, rights management, hosting, metadata management, indexing - Accessibility is guaranteed, easy to find, long-term available - Added to libraries - Print versions - Influence on academic ranking/impact factors - May be required by funding agency And who pays publishing fees? - Depends on license (publisher, author, or community) - Community: Conference participants, tax payers (in Germany: ["Project DEAL"](https://www.projekt-deal.de/about-deal/)) Attention: There are [*predatory publishers*](https://en.wikipedia.org/wiki/Predatory_publishing)! ## Paper Repositories Non-authorative versions: - [Homepage](https://www.sosy-lab.org/research/bib/All/index.html) - [ArXiv](https://arxiv.org/) Authorative-versions (also called 'version of record'): - at publisher repositories: Springer Publishing, ACM, IEEE, Elsevier, Wiley, etc. (These organize conference proceedings and journals) - Popularity and use of individual repositories depends on concrete science. ## Bibliographic Data - [Homepage](https://www.sosy-lab.org/research/bib/All/index.html): Low confidence. May be incomplete or inconsistent -> Not suited for citation - [DBLP](https://dblp.dagstuhl.de/): Highest confidence. Human-maintained and reliable. - [Google Scholar](https://scholar.google.de/): Complete, but automatically generated and may be inaccurate (e.g., uses 'et al.'). But: Nice notifications about new papers that cite own work - [ORCID](https://orcid.org/): Automatically generated, grows in use. Nice notifications about new *own* papers, connected via digital object identifiers (DOIs). - [CrossRef](https://www.crossref.org/) (papers) and [DataCite](https://datacite.org/) (data): DOI agencies, manage metadata ## Ideal Requirements for Publishing Hard requirements: - long-term available (archived) - immutable (is always the same version referenced?) - identifiable (is the correct version referenced?) DOI is industry standard for this. Weak requirement: Open Access ## Unique Identification - DOI: Digital Object Identifier, mostly for papers - Resolver: DOI -> URL (e.g., doi.org) - points to landing page of publication - linked to metadata - DOI pointers to landing-page URLs are maintained - ORCID for scientists - Archival Resource Keys ([ARK](https://arks.org/)) for all kinds of objects # Publications: Details *This chapter is based on the notes [2022-05-17_SPSE_Publication-Part2.pdf](../../course_notes/2022-05-17_SPSE_Publication-Part2.pdf)* ## Open Access and Open Science - People **always** have been asking for **access** to knowledge - But access is limited through society and technology. Examples for knowledge transfer: ## Paper Preprint - paper version accepted at conference - not the officially published version - part of green open access ## Contributing to Software: - Contributor-License-Agreement (CLA) * maintainers free to use, share and modify * legally sound - no CLA * simple process, no entry barrier * no confirmation by contributor's employer necessary - [Developer Certificate of Origin (DCO)](https://opensource.com/article/18/3/cla-vs-dco-whats-difference) * contributor asserts that she is allowed to commit the code * clarifies license of the committed code patch ## Making copyright and licenses explicit Very important! [SPDX](https://spdx.dev/) is common format to share important data - clarifies how software (patches) are licensed - easy to maintain - automated checks with [REUSE](https://reuse.software/) Example for Java: [CPAchecker Source Code](https://gitlab.com/sosy-lab/software/cpachecker/-/blob/349bc23f03af7941a9d7fa09e567fcb54dcf3f7c/src/org/sosy_lab/cpachecker/core/CPAchecker.java#L1-7) ## Artifacts Reproduction packages. ([our Wiki page on the topic](https://gitlab.com/sosy-lab/doc/-/wikis/Reproduction-Artifacts)) Examples: - Data sets, statistics - software - machine-readable proofs - appendices (Don't forget to license!) Three dimensions: - Repeat - Reproduce - Replicate ACM badges qualify papers with regards to their reproduction packages: - Artifact available - Artifact evaluated (a) functional and/or (b) reusable - Results reproduced - Results replicated Repositories: - [Zenodo](https://zenodo.org/) - [FigShare](https://figshare.com/) - [Code Ocean](https://codeocean.com/) - [Mendeley Data](https://data.mendeley.com/) ### Discussions * Is replication always good? Example: [Google](https://www.reuters.com/technology/google-faces-internal-battle-over-research-ai-speed-chip-design-2022-05-03/) * Negative reproduction study example: https://dl.acm.org/doi/pdf/10.1145/3340571 ## Publication Process *This chapter is based on the notes [2022-05-24_SPSE_Rankings.pdf](../../course_notes/2022-05-24_SPSE_Rankings.pdf)* ![](3-Publications/Process.png)