A table with duplicated rows. Duplicated rows only containing one unique property and except of the identifier all other properties are empty.

• Comma separating image links into one column
• Removing duplicate columns

Lets start to define records in the system. The first column of the table is the identifier of a record. All records with an empty first column are associated to the first record above with a value in the first column.

Sort

In OpenRefine, sorting is only a visual aid, unless you make the order permanent. To do this, click the Sort menu at the top and choose Reorder rows permanently.

• Move Identifier column to start of the table (Edit column > Move column to beginning)
• Sort by first (Identifier) column (Sort > by text)
• Sort by Property (Sort > by text)
• Reorder permanently (option in the top panel of the table: Sort > Reorder rows permanently)

Interim Result

Define Records

• Remove duplicated Identifiers (Edit cells > Blank down)
• Switch to records in table options. Now you should see your records divided by color coded rows

• Join cells of the Image-Link column (Edit cells > Join multi-valued cells)
• Define separator

Interim Result

• Use facet to select all empty rows on the first column (Facet > Customized facets > Facet by blank)
• Switch to row view in the top panel of the table
• In the Facet/Filter on the left include true
• Now you should only see the rows you like to remove
• Remove all rows in the All row (Edit rows > Remove all matching rows)
• Remove Facet/Filter by clicking the Remove All button

Result