Review Comment:
The manuscript presents a very interesting tool for creating storytelling maps in an easy way. The work is an extension of their existing tool NBVT, where a spatial dimension of the knowledge visualization is added by introducing storytelling maps. The demo presented very nice results, which could potentially be an attractive application for many users and agencies. Therefore, it could be interesting work for the cartography and gis communities as well.
The following are suggestions that the authors may consider for improvement:
1. Links to Wikidata are automatically generated, which are interesting for further reading. In many cases they are too many, so people may ignore them, or get lost within them. It would be great (in my mind) if the entities in the texts could have hyperlinks, which is similar to Wikipedia's layout.
2. The evaluation of entity extraction performance is not reported. Maybe the authors have reported it in their early publications, but still, it could be nice to have it mentioned in this manuscript that tells how well the automatic process worked.
3. Sec.3: Five participants for evaluating SMBVT could be not very convincing.
4. Detailed instruction on how to deploy the system is currently missing from the git repo (https://github.com/EmanueleLenzi92/SMBVT).
5. License is missing as it is mentioned earlier to be open source in the early published version (https://ceur-ws.org/Vol-3370/paper18.pdf).
6. To comply with the "long-term repository discoverability", the repo could be released/published via Zenodo to obtain doi.
7. Sec. 4 The limitations of the tools are less discussed in this chapter.
8. There would be a necessity to demonstrate the efficiency/ performance of the online tool with metrics that are often used to evaluate web systems.
Some minor ones:
1. P2L37 double points
2. P5L45 Which of the popular software would be supported to use these json files directly, some examples would be very helpful.
|