This algorithm merges all people in an ISI database who have the same name (absent case and punctuation).
Data Preparation > Database > ISI > Merge Identical ISI People
The merging is performed as indicated in Merge Entities. The algorithm merges on the RAW_NAME column from the People Table table. This column has the short form of a person's name (last name, first initial and maybe middle initial). To identify identical entities, this algorithm compares all the "normalized" values from the RAW_NAME. In this case, "normalized" means that the value has been converted to lower case and all characters that are not a decimal, letter, or a single space are removed. If the "normalized" values are the same, the people being compared are assumed to be identical. There is some limited 'sanity checking' that checks if the people being merged were authors on the same paper with the assumption that this case should not occur. If it finds such a merge group, it will warn the user on the console with the RAW_NAME, primary key, ISI_UNIQUE_ARTICLE_IDENTIFIER, and primary key but will continue with the merge.