Chatbot-Based Natural Language Interfaces for Data Visualisation: A Scoping Review

Round 1

Reviewer 1 Report

Following are some comments to be addressed before publication:

1. Authors should add contribution of the paper at the end of Introduction in bullets form.

2. The authors have used PRISMA as reporting framework for SLR. Can this framework address the issues relating to Publication Bias, Language Bias, updating the review, reporting the funding sources and conflict of interest, etc.

3. There is not enough  materials available to evaluate the general and specifically chatbot user interfaces. There are established guidelines of HCI for interfaces evaluation. Authors should add description from the following references which should be cited in the paper. 

i. Khan, I., & Khusro, S. (2020). Towards the design of context-aware adaptive user interfaces to minimize drivers’ distractions. Mobile Information Systems, 2020.

ii. Khan, I., & Khusro, S. (2022). ConTEXT: context-aware adaptive SMS client for drivers to reduce risky driving behaviors. Soft Computing, 26(16), 7623-7640.

English is fine only minor edits need to be done while submitting the revision.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Dear authors,

I carefully read your article, which performs a very detailed bibliographical review relating Chatbot with Visualization-oriented Natural Language Interfaces (V-NLI). To understand this article it is necessary to emerge in the technology developed since the beginning of this century. It should be understood that creating correct visualizations to help users easily understand data is quite a complex task from an Artificial Intelligence point of view. This is a very detailed and very interesting survey. Reading this article is enjoyable from beginning to end.

The Introduction was very well written, explaining how the first computer systems (used for data visualization) evolved into the famous Chatbots. Furthermore, at the beginning of Section 2, Figure 1 is quite enlightening and didactic to understand the entire article. The authors also make a detailed description between Data Visualization (Section 2.1) and Chatbot (Section 2.2). Also in Section 2.2, the authors propose a general characterization of Chatbots using 4 dimensions called AINT (see Figure 2). Section 5 provides 8 Remarks that will be very helpful for researchers to develop and improve the systems described in this article.

This article was one of the best articles I reviewed for MDPI. The article is well-written and well-organized from beginning to end. I am also in favor of its publication in MDPI's Journal of Applied Science. However, I think I can help, the authors of this article, with some small improvements of the same. So here are my thoughts:

1. What are the abbreviations AR and VR at the end of the abstract? Make this clear to the reader who is not from the area. You also mention this acronym in Section 2.2 (p. 10 and lines 391 and 392), but again, you say nothing of its meaning. They appear again in Section 5.

2. What I am going to say in this paragraph is just a suggestion (not mandatory). Since data visualization is applicable in any branch of human knowledge (e.g., science, education, engineering, medicine, etc.) would not it be interesting to change the keyword “Data Visualization” for something like “World's Data Visualization” or “World Data Visualization”? I think this would increase the visibility of the article.

3. In the caption of Figure 3 (p. 7) a comma is missing before the “and”.

4. Is there an “and” missing in the caption of Figure 4 (p. 8) before item (e)? Just check this!

5. I think there is also something wrong with the numbering used in the caption of Figure 5 (p. 8). There is an extra “and” (before the number (3)) and the number “(2)” should be on the left of the word “visualization” and not on the right. Just check this!

6. In Keywords, separate each term with a semicolon and not a comma. If Keywords are neither proper names nor acronyms, it would be better to start with a lowercase letter.

7. (p. 10 and line 395) Replace “the the PRISMA-ScR...” with “the PRISMA-ScR...”.

8. I made a visual checklist in the numerical order of the citations. It is OK!

9. In the caption of Table 2 (p. 14) replace “...Rule-based and Intelligent” with “...Rule-based, and Intelligent”.

10. (p. 14 and 15) I think lines 518, 545, and 557 are missing a comma. Only check these lines!

11. (p. 19) In Table 4, the cell titled “C. Guidance” would not it rather refer simultaneously to the cells “Help”, “Autocom.”, and “Recom.” and not just the “Help” cell (as it is now)?

12. (p. 20 and line 740) Replace “...to natural language (NL).” by “...to Natural Language (NL).” to highlight the abbreviation.

13. You cite Table 5 only on page 24, but it appears - well before that - on page 21. Would not it be better to start Section 4.5 (p. 20) already citing Table 5?

14. (p. 21 and line 777) Change “... by date.), ...” to “... by date), ...”. It has an unnecessary full stop.

15. In Section 5 there are “8 Remarks” and not just 7! There are two Remarks with the same number (Remark 5).

16. At the end of the Conclusion (line 1074) Replace “VNLI” with “V-NLI”. It is always desirable to maintain uniformity of acronyms throughout the text.

17. The Bibliographical References appearing on pages 28 through 33 are slightly outside MDPI Journals standards. All words in the name of the Journal, of a given reference, begin with a capital letter and are in italics. You highlighted the names of the Journals (this cannot be done) and some of them are starting with lowercase letters. Also, in general, the DOI of the article is placed at the end of each reference. However, I do not know if the DOI is mandatory. Personally, I do not care for the DOI to appear.

Note about the English of the article: in my opinion, the article was very well written in the English language. However, as I am not a native speaker of this language, I will not be able to help you with small mistakes, which only a profound connoisseur of the English language would be able to make.

Yours sincerely,

 

The Reviewer.  

Note about the English of the article: in my opinion, the article was very well written in the English language. However, as I am not a native speaker of this language, I will not be able to help you with small mistakes, which only a profound connoisseur of the English language would be able to make.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

In this manuscript, the authors have performed a scoping review to analyze the synergies between the fields of Data Visualisation and Natural Language Interaction. The focus was on chatbot-based Visualisation-oriented Natural Language Interfaces (V-NLI) approaches. Research findings show that the works in the literature put a strong focus on exploring tabular data with basic visualizations, with visual mapping primarily reliant on fixed layouts. Moreover, V-NLIs provide users with restricted guidance strategies, and few of them support high-level and follow-up queries. The research is well-done and presented. The discussion is well written. The results, outlined research gaps, and future research directions are promising in advancing this field.  

 

I have the following comments.

 

1. Line 122: “…Since the input data is in a tabular format…” Could you explain more about why the input data is restricted to a tabular format?

 

2. Line 161: “The Visual Mapping Step involves defining: the spatial substrate - i.e. the space and the layout used to map the data - the graphical elements - i.e. marks such as points, lines, images, glyphs, lines, etc. - and the graphical properties - also called retinal properties, i.e. size, colour, orientation, etc.”

I suggest listing the items in a numbered list for improved readability (and to avoid such a long sentence).

 

3. Line 218: “…we use the following categories: fixed, user-defined, rule-based, and intelligent methods.”

Rule-based methods/systems can be advanced (with some sort of intelligence) and, in that case, considered “intelligent methods” (e.g., like in rule-based expert systems). Please elaborate on why you consider “rule-based” a category different from “intelligent methods.”

 

4. 1. Figure 1, under “Data Transformation” box, please change bining to binning.

 

5. Line 396: “…introduced by [78].” I recommend improving this to: “…introduced by Tricco et al. [78].”   

 

 

6. Line 423: “(iii) Articles that did not present and contribute original work.” Could you please explain this point further (e.g., How did you decide on this? (this is a subjective criterion (not straightforward); how many papers were excluded because of this criterion? etc.).

Thank you.

Moderate editing of the English language is required.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

The authors incorporated all my concerns. Now accepted from my side.

English is fine.