The main difference between A temporal dataset and normal dataset is time. Let's look at this way, when every single record (row) in the dataset has a timestamp (date, date+time) associated with it, then this dataset could be considered for temporal analysis e.g. forecasting. Of course we associate a timestamp to each row using a temporal column (e.g. time, date, timestamp). I looked at the 'Pima Indians Diabetes Database' on Kaggle but I could not see any associated timestamp information so I would consider this as a normal dataset.

Not always a temporal dataset is used for temporal analysis. It always depends upon the use case. E.g. if we have a loan dataset which has temporal information on the borrowers. One can for example classify these borrowers into fraud or non-fraud categories using a logistic regression -> no time information is needed here or one can also do a survival analysis to find who will default the loan and likely when. But the key is that we know what the problem we want to solve given a temporal/normal dataset. (although second paragraph is not needed but I just added in case it clarifies further)

Hope this helps!

Dear @Nasif Muslim

The main difference between the temporal and non-temporal data is a time constraint is appended with data representing when it is applicable or stored in the database. Temporal data is created by time-stamping the normal data.

Without going into implementational details, one can imagine that - whenever an update occurs - a conventional database physically updates, i.e. throws old values away and stores the new ones, whereas a temporal database is updated logically, i.e. it marks the old and new values with timestamps that indicate to which.

Temporal Databases – A Review

https://www.google.com/url?sa=t&source=web&rct=j&url=http://ijrar.com/upload_issue/ijrar_issue_1060.pdf&ved=2ahUKEwiqxomxhdz1AhXLQfUHHQGrDl4QFnoECAcQBg&usg=AOvVaw0qvvS3gCWLA4Wu4JCTZfgt