10 KiB
10 KiB
type:: REVIEWS tags:: year:: 2024 venue:: MODELS-DS full-title:: Estimation, Impact and Visualization of Schema Evolution in Graph Databases date-start:: 22-07-2024 - 16:12 date-submitted:: external-links:: status:: DONE deadline-submission:: file:: @Estimation, Impact and Visualization of Schema Evolution in Graph Databases parent:: todoist:: https://app.todoist.com/app/task/2-dominique-hausler-estimation-impact-and-visualization-of-schema-evolution-in-g-6Vm6Gv2GqXmhhrM6
- ### [[Highlights]]
-
- # Annotazioni
- (22/7/2024, 16:05:35)
- - “Schema Evolution in Graph Databases” ([Hausler, 2024, p. 1](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=1&annotation=QWWKJ7D5)) #5fb236
- - “Due to NoSQL databases being schemaless, heterogeneous data can occur when performing data changes.” ([Hausler, 2024, p. 1](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=1&annotation=572G7U9W)) #5fb236
- - “As a user can not foresee the results of the evolution operation neither the amount of data changes nor the possible schema violations or a relaxed schema, a system to show the impact of evolution is essential.” ([Hausler, 2024, p. 1](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=1&annotation=VQCPHW5A)) #a28ae5
- - “estimate and illustrate the impact of evolution on the schema level” ([Hausler, 2024, p. 1](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=1&annotation=SWQ9939G)) #5fb236
- - “structure profiles are designed for an initial data exploration” ([Hausler, 2024, p. 1](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=1&annotation=RSWRHA75)) #5fb236
- - “Due to graph database offering schemalessness, heterogeneity can occur through optional elements or structural error.” ([Hausler, 2024, p. 1](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=1&annotation=IEUBAJSS)) #5fb236
- - “our goal is to involve users by informing them about such occurrences and giving the option to make changes if needed.” ([Hausler, 2024, p. 1](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=1&annotation=D78JWW8P)) #5fb236
- - “Schema Management Layer (SML)” ([Hausler, 2024, p. 1](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=1&annotation=Q3HBL7VM)) #5fb236
- - “schema and statistical information can be extracted, compared and constraints detected.” ([Hausler, 2024, p. 1](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=1&annotation=MQP4HZ7X)) #ffd400
- *This seems not new. What's the novelty with respect to existing work?
- I'm missing relevant work from Atzeni et al. like:
- Paolo Atzeni, Francesca Bugiotti, Luca Cabibbo, Riccardo Torlone: Data modeling in the NoSQL world. Comput. Stand. Interfaces 67 (2020)
- Paolo Atzeni, Francesca Bugiotti, Luca Rossi: Uniform access to NoSQL systems. Inf. Syst. 43: 117-133 (2014)
- Francesca Bugiotti, Luca Cabibbo, Paolo Atzeni, Riccardo Torlone: Database Design for NoSQL Systems. ER 2014: 223-231
- Concerning the evolution management, the author should also consider:
- Maxime Gobert, Loup Meurice, Anthony Cleve: Modeling, manipulating and evolving hybrid polystores with HyDRa. Sci. Comput. Program. 230: 102972 (2023)
- Maxime Gobert, Csaba Nagy, Henrique Rocha, Serge Demeyer, Anthony Cleve: Best practices of testing database manipulation code. Inf. Syst. 111: 102105 (2023)
- Anthony Cleve, Maxime Gobert, Loup Meurice, Jerome Maes, Jens H. Weber: Understanding database schema evolution: A case study. Sci. Comput. Program. 97: 113-121 (2015)
- Compared to such related work, I see potential novelty mainly concerning RQ3 and RQ4.*
- - “the schema extraction paired with the information extraction for structure profiles is new” ([Hausler, 2024, p. 1](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=1&annotation=E5Q85EZV)) #ffd400
- *Why? What's the limitation of existing work?*
- - “RQ3: How to estimate the effort and illustrate the impact of evolution operations as well as the affected elements? • RQ4: What kind of optimizations are needed?” ([Hausler, 2024, p. 2](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=2&annotation=YVVH5WBQ)) #ffd400
- *Maybe this part is new. Concerning RQ1 and RQ2, they are related to existing work that is not mentioned.*
- - “his limitation is due to schema evolution affecting only the property names np .” ([Hausler, 2024, p. 3](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=3&annotation=4WYVHYT4)) #5fb236
- - “estimate and display the affected elements through the Preview Module.” ([Hausler, 2024, p. 4](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=4&annotation=ZD6HEHMD)) #ffd400
- *What do you mean with estimate? What kinds of estimation are you supporting?*
- - “Schema evolution is conducted through the evolution operations add, rename, delete, copy, move, split and merge” ([Hausler, 2024, p. 5](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=5&annotation=6QA66XMC)) #5fb236
- - “estimation of the impact of evolution on the database” ([Hausler, 2024, p. 5](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=5&annotation=BY963PWJ)) #5fb236
- - “easy-to-understand language called GEO.” ([Hausler, 2024, p. 5](zotero://select/library/items/4PIZV7JR)) ([pdf](zotero://open-pdf/library/items/9ZGTXUZS?page=5&annotation=Y8IJGKSZ)) #a28ae5
- ### [[Comments]]
-
- The paper presents an approach to managing schema evolution in graph databases. The paper is about a relevant problem even though it partially compares with existing literature and thus requires a more detailed justification of its novel contributions. In the following, the paper is discussed with respect to its value, impact, originality, rationale, research plan, and overall quality.
- Value: The problem addressed in this paper is highly relevant given the increasing use of graph databases in various domains. The problem is worth exploring as it has significant implications related to the challenges of management and evolution of large-scale data systems.
- Impact: By providing a framework to estimate, illustrate, and manage the impact of schema evolutions in graph databases, the proposed approach can improve current practices in database management. I see the potential impact, especially concerning the estimation of evolution operations and their optimization (i.e., RQ3, and RQ4).
- Originality: While the paper proposes novel ideas, it must adequately distinguish its contributions from existing work. The paper needs a thorough comparison with key related works, such as those by Atzeni et al. and Gobert et al., which address similar issues in NoSQL databases. For example, works such as "Data modeling in the NoSQL world" by Atzeni et al. (2020), "Uniform access to NoSQL systems" by Atzeni et al. (2014), and "Modeling, manipulating and evolving hybrid polystores with HyDRa" by Gobert et al. (2023) provide relevant insights and methods that should be considered. Including these comparisons and clearly articulating the novelty of the proposed solutions would strengthen the paper's claim to originality.
- Rationale: The paper provides a sound rationale for the proposed approach, highlighting the necessity of schema management in maintaining data integrity in graph databases. However, the paper could benefit from a more detailed discussion of the limitations of existing methods and how the proposed solutions specifically address these gaps. Moreover, even though the estimation part of the work is very interesting, the paper does not properly discuss the quantitative instruments that users can rely on to perform the estimation of evolution operations.
- Research Plan: The research plan outlined in the paper appears realistic and well-structured. Integrating schema management, structure profiles, and a preview module into the Nautilus tool is well-conceived. However, the plan could be improved by more explicitly detailing the evaluation metrics and methodologies, particularly concerning the estimation accuracy and impact assessment of schema changes.
- Quality: Overall, the paper is of good quality, and the writing is clear. However, the related work section needs to be more comprehensive, including comparisons with significant related research. Relevant works that should be considered include:
- Atzeni, Paolo, Francesca Bugiotti, Luca Cabibbo, Riccardo Torlone. "Data modeling in the NoSQL world." Comput. Stand. Interfaces 67 (2020).
- Atzeni, Paolo, Francesca Bugiotti, Luca Rossi. "Uniform access to NoSQL systems." Inf. Syst. 43: 117-133 (2014).
- Bugiotti, Francesca, Luca Cabibbo, Paolo Atzeni, Riccardo Torlone. "Database Design for NoSQL Systems." ER 2014: 223-231.
- Gobert, Maxime, Loup Meurice, Anthony Cleve. "Modeling, manipulating and evolving hybrid polystores with HyDRa." Sci. Comput. Program. 230: 102972 (2023).
- Gobert, Maxime, Csaba Nagy, Henrique Rocha, Serge Demeyer, Anthony Cleve. "Best practices of testing database manipulation code." Inf. Syst. 111: 102105 (2023).
- Cleve, Anthony, Maxime Gobert, Loup Meurice, Jerome Maes, Jens H. Weber. "Understanding database schema evolution: A case study." Sci. Comput. Program. 97: 113-121 (2015).
- ### [[REVIEWS/Notes]]
- ### YELLOW CONCERNS
background-color:: yellow
- {{query (and [[ffd400]] [[DocSymMod24_2]] )}}
collapsed:: true
- ### ❓️Questions
- {{query (and [[question]] [[DocSymMod24_2]] )[[question]]}}
query-table:: true
query-properties:: [:block]