Accommodating substructure similarity-based search in a recipe database system

Improving the precision of information retrieval has been a challenging issue on Chinese Web. As exemplified by Chinese recipes on the Web, it is not easy for people to use keywords (eg, recipe names) to search recipes, since the names of Chinese dishes can be literally so abstract that they do not bear much, if any, information on the underlying ingredients or cooking methods. In this paper, we investigate the underlying features of Chinese recipes, and based on workflow-like cooking processes, we model recipes as graphs. Benefiting from the characteristics of graphs, we mine frequent common patterns in a cooking graph database. We also propose a novel similarity measure based on the frequent patterns, and devise a novel filtering algorithm to prune unrelated data so as to support efficient and effective on-line searching. © 2008 IEEE.