A Cloud-Native Web Application for Assisted Metadata Generation and Retrieval: THESPIAN-NER^†

Round 1

Reviewer 1 Report

The purpose of this research is to propose a cloud-native web application that can assist in the development and retrieval of metadata. In addition to designing, developing, and training the models, the authors examine the design of the web app and the experiences that users have with it. In the end, this work was incorporated into the platform-as-a-Service that is hosted in the cloud by 4CH. However, in the end, the work is written effectively and is nicely organized. The goals or contributions of this study should be included in the introduction section.  Please resolve line 146 of the paper's structure.

Author Response

Dear Reviewer#1,

Thank  you  for  giving  us  the  opportunity  to  submit  a  revised  draft  of  our  manuscript  titled “A cloud-native web application for assisted metadata generation and retrieval: THESPIAN-NER” to the special issue if MDPI/Computing and Artificial Intelligence Recent Trends in Natural Language Processing and Its Applications.

We appreciate the time and effort that you dedicated to providing your valuable feedback. 
We have been able to introduce changes to reflect the provided suggestions. 

A point-by-point response to the comments is provided.

Sincerely,
Alessandro Costantini, on behalf of the Authors

Author Response File: Author Response.pdf

Reviewer 2 Report

The work is a nice example of engineering and ensamble of the current most advanced technologies in web services and NLP.

The web application is designed for management of metadata using a NER module. There is no novelty in any of the parts of the design. The most advanced technique is the ensambling of advanced technologies to achieve the THESPIAN-NER system.

The authors report some experiments of classification for ArcheoNER and hsNER based on CNN and Transformers. Both methods achieve very poor results, especially in the ArcheoNER model. It is difficult to sustain an application with this results. The authors plan to improve the results and they give some (vague) clues on how to do that. They should elaborate the perspectives of future in this key point.

I understand that the application is only able to deal with Italian texts. This is a European Project with European funding and an ambitious objectives. I wonder if there exist any plans of making the system available to be used with other European languages.

Author Response

Dear Reviewer#2,

Thank  you  for  giving  us  the  opportunity  to  submit  a  revised  draft  of  our  manuscript  titled “A cloud-native web application for assisted metadata generation and retrieval: THESPIAN-NER” to the special issue if MDPI/Computing and Artificial Intelligence Recent Trends in Natural Language Processing and Its Applications.

We appreciate the time and effort that you dedicated to providing your valuable feedback. 
We have been able to introduce changes to reflect the provided suggestions. 

A point-by-point response to the comments is provided.

Sincerely,
Alessandro Costantini, on behalf of the Authors

Author Response File: Author Response.pdf

Reviewer 3 Report

From what I gather in the abstract, there's a project called Competence Centre for the Conservation of Cultural Heritage (4CH) and within the project, European competence centre of Cultural Heritage is envisioned and information platform is established to help digitalize storage of research work (papers, data...). In order to help the process of data collection, a service is developed that enables automatic recognition of named entities in uploaded documents. And then authors deal with now to choose the deep learning model to perform the recognition, use it for the web service and integrate into 4CH framework, which is what I find the most important contribution.

The structure of the paper is somewhat hard to follow, English language needs a lot of improvement and proofreading. The paper would benefit from one person going through the entire text and remove duplicates and consolidate it. Eg.: FAIR principles are first mentioned in line #24, then the abbreviation is explained in #46 and again in #72.

Why was JSON/MongoDB selected? Did you try other databases?

A note: significant part of the research was already published in https://link.springer.com/chapter/10.1007/978-3-031-13324-4_23, which is not referenced in this work. 

Author Response

Dear Reviewer#3,

Thank  you  for  giving  us  the  opportunity  to  submit  a  revised  draft  of  our  manuscript  titled “A cloud-native web application for assisted metadata generation and retrieval: THESPIAN-NER” to the special issue if MDPI/Computing and Artificial Intelligence Recent Trends in Natural Language Processing and Its Applications.

We appreciate the time and effort that you dedicated to providing your valuable feedback. 
We have been able to introduce changes to reflect the provided suggestions. 

A point-by-point response to the comments is provided.

Sincerely,
Alessandro Costantini, on behalf of the Authors

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

Thank you for your reply to my initial review. Good luck!