Finding adjacent words in documents means that the database must
maintain the exact position of each occurrence of a word in a document.
Combining substring and adjacency are not well defined.
To find something like "act of life", the problems are:
- The system my have ignored the stopword "of", and thus may match
"act in life" as well.
- The system may only know positions, and by ignoring the stopword
may match "actual life".
- Should the system match each word as substring or not?
The user may wish to find "facts of life".
