Table 11 Types of textual patterns and clues for identification of database and software names
From: Ambiguity and variability of database and software names in bioinformatics
Type | Contribution to total TPs |
|---|---|
Machine learning matches | 55.3 % |
Heads and Hearst Patterns | 9.8 % |
Title appearances | 0.5 % |
References and URLs | 1.8 % |
Version information | 0.9 % |
Noun/verb associations | 21.4 % |
Comparisons | 4.0 % |
Remaining | 6.3 % |