Database query problems can be categorized by the expressiveness of their query languages, and data structure bounds are better for less expressive languages. Highly expressive languages, such as those permitting Boolean operations, lead to difficult query problems with poor bounds, and high dimensionality in geometric problems also causes their query languages to become expressive and inefficient. The IDSgrep kanji dictionary software approaches a highly expressive tree-matching query problem with a filtering technique set in 128-bit Hamming space. It can be a model for other highly expressive query languages. We suggest improvements to bit vector filtering of general applicability, and evaluate them in the context of IDSgrep.
|Titel||Similarity Search and Applications: 9th International Conference, SISAP 2016, Tokyo, Japan, October 24-26, 2016|
|Status||Udgivet - 2016|
|Navn||Lecture Notes in Computer Science|