5.4 Similar codes browse list. The system constructs lists of spellings that are similar to a certan name by finding all the spellings thaving a code that up to a point is the same as a spelling in its group. As the group corresponding to the focus in the lexical browse changes, the contents of this window changes. There are eight grouping principles whose codes make sense as an ordering principle on similarity: spelling, normal, viewex, Soundex, NYSIIS, phonetic, dialect, and compressed. If two spellings have the same code, these two spellings belong to the same group according to the grouping principle producing that code. One definition of similarity means that the code is identical only up to a certain point. At present we use the rule that 5/6 of the contiguous characters, either from the beginning of the string or from the end, are identical. This means that code strings up to five characters must be identical. This kind of partial identity has the effect of bringing in many more spellings for consideration in makeing a grouping decision. Each spelling similar to a member of the group being displayed has an associated set of indicators for the code(s) its similarity is based on.