Predicting the Effectiveness of Keyword Queries on Databases

Keyword query interfaces (KQIs) for databases provide easy access to data, but often su er from low ranking quality, i.e. low precision and/or recall, as shown in recent bench- marks. It would be useful to be able to identify queries that are likely to have low ranking quality to improve the user satisfaction. For instance, the system may suggest to the user alternative queries for such hard queries. In this paper, we analyze the characteristics of hard queries and propose a novel framework to measure the degree of di culty for a key- word query over a database, considering both the structure and the content of the database and the query results. We devise e cient algorithms to compute the degree of di culty at query-time, and show that the overhead is very small com- pared to the query execution time. We evaluate our query di culty prediction model against two relevance judgment benchmarks for keyword search on databases, INEX and SemSearch. Our study shows that our model predicts the hard queries with high accuracy.published or submitted for publicationnot peer reviewe