Abstract: The catalogue retrieval systems in water sector are good support for field users, but sometimes are difficult to start for ordinary users without clear needs, and also are easy to produce information overload problem. To this end, the faceted search technology is introduced into water conservancy, and a faceted recommendation algorithm based on discrimination is proposed. In this algorithm, the search path is constructed as a search tree which can ensure the user’s search path is the shortest. To prove the feasibility and validity of the method, the algorithm is applied to the actual retrieval system, and the water conservancy object is set up and classified, so that different users can retrieve water data from different dimensions. The new system greatly improves the efficiency of user’s retrieval, and provides a new perspective for the water conservancy application system.

Key words: classified label, dipartite degree, facet search