INTRODUCTION

• The Neotoma database is used by researchers studying ecological changes over the past 5 … years. “used “is not the best verb in relation to a database. What do you think?. This sentence could be improved.
• Comprehend ecological changes comprehensively
  Not the best phrasing
• “This will be done in three parts”
  I wouldn’t use the future tense, you have already done it! This is not a proposal report.
  DATA SCIENCE METHODS
  2.1
• “The model will predict”
  “The model is able to predict” or another similar rephrasing not in the future tense.
• Add a caption to Table 1 and refer to it in the text.
• With the help of our partner Simon, a list of articles that currently contribute to Neotoma was provided for the positive cases
  This is a technical report. Can you describe objectively this? For example, what percentage of the Neotoma database represents this list of articles? What criteria has Simon used to select the articles? Were they randomly selected or you were given a representative sample?
• I like Table 2. You should increase the width of the first column to improve the readability.
• Author & Journal Subject - Why you took this decision? The sentence it is a bit vague.
• The table is called Feature Selection Desitions. It is difficult for me to recognize the different rows of the table… For example, “Author & Journal Subject” seems to have 2 rows associated. I am not sure that I am understanding well the table, could be better formatted.
  2.1.1.2
  CrossRef API.
• All the databases that you extract information from should be cited. Also you should include the day you access the information in the reference section to increase the reproducibility.
• Table 2 is not cited in the text. “The summary table below should be replaced by “Table 2”.
  2.1.1.3
  Several text representations experimented
• Include what did you try and why you decide to keep these n methods. List the all the methods.
• Adding term-association probability with zero-shot classification “This method was not used eventually”. Are you listing the methods you use or the ones you didn’t use?. If you didn’t use it, you can name it as something you tried, but I would not mix it with the other methods as it is unclear for the reader.
• Table 3 is not mentioned in the text.
  2.1.1.4
• ” A variety of SML were experimented”
  “We explored/experimented with the models A, B and C”. Then you can go one by one. I would use bullet points to reduce the text, not only when listing the types of models if not also listing the models themselves. There some repeated phrases in this section of the text.
  2.1.2

2.2 Ty completing

2.3
I would add links to the app here and also in other sections of your text.
2.3.1

• I would just add why you decide to use Dash. It is why a potential reader could be interested on.
  3.1
• When you discuss results cite the figure and you can include the percentages for better analysis of the text
  For example “Naive Bayes has the highest recall score (96.7%) but the precision was low (70.1%) and will introduce false positives (Fig. 1).
• If you have many results that could be interpreted from the same figure you can cite it at the end of that paragraph.
  There are problems with the captions in tables and figures in this section. There aren’ rendering.
  3.4
• “It will be scheduled to run on a daily basis”
  It is expected to run on a daily basis. / The plan is to run it on a daily basis
• Be careful with long paragraphs. Each paragraph should have one concrete meaning.
• The diagram should be a figure and should be cited in the text.
  4