Understanding full.dataTables.html in ami search

ami-search results

The results of ami-search are collected in full.dataTables.html.

raw tables

Here's a typical example from a search for "N95" (a type of face mask) (note the URL prefix to make the file display). The table has over 352 rows so a long way to scroll down.

https://htmlpreview.github.io/?https://github.com/petermr/openVirus/blob/master/searches/n95/full.dataTables.html

details

assets This is a 5 * 352 table. The rows are articles (Hits) from the search. There are 7 columns:

1. Unique Id (CTree name) and links to remote (EuropePMC) and local (scholarly.html) copies of the article. (always present).
2. Bibliography (always present). Title, date (bug), journal (bug) and abstract. To get values mouse-hover.
3. dictionaries in alpha order (depends on search). Here we used the builtin dictionaries "country" (from Wikipedia), "disease" (from Wikipedia, based on ICD), "funders" (more generally organizations, from Crossref), and "medproc" (generated by us from Wikipedia page on "Medical Procedure".
4. word frequencies. Note for n dictionaries there are n+3 columns.

We use stemming (e.g. "hospital~" for "~s" , "~ised") but not synonyms. Cells 3... contain the most frequent hits/words. Numbers represent the frequencies. The first article had 59 mentions of the word "patient" or "patients" . All words are linked to Wikipedia pages or disambiguation pages. Sometimes we guess wrong!

dataTables

The file contains Javascript which links to the Open "dataTables" tool. This colours the columns, creates pages and tabbing.