Interfaces that are not user-friendly often limit the usability of a system. The author notes that user interfaces that do not take human factors into consideration frustrate users and limit the functionality of the system.

Query interfaces for database systems support keyword-based filters. For example, in Amazon’s user interface, a user could filter books based on their prices ($0 to $10, $10 to $50, and so on). However, the criteria for filtering the results of a query rely on an exact match between the supplied keywords and the entries in a database table, which the author notes as too stringent a requirement for users, either due to an “incomplete memory” or due to typographical errors. To cope with users’ tendencies to specify incomplete keywords, the author proposes that query filters support partial matches. To support partial matches, the equality conditions in a query are replaced by regular expressions that include wild card characters.

The article is motivated by the fact that humans often either (a) fail to specify the correct keywords, or (b) specify incomplete keywords. The author’s proposal solves only a small fraction of the problem. By replacing equality conditions in a query with wild card expressions, only the second case (b) is solved. However, allowing for partial matches doesn’t help users query a database system when misspelled or incorrect keywords are specified. That said, the author’s proposal could be generalized to handle misspellings by defining a notion of similarity/dissimilarity between words [1,2].

Although the article addresses a very small fraction of the bigger problem of designing more user-friendly query interfaces, it encourages readers to focus on redesigning query interfaces to expose the querying functionalities in a more “humanized” way.