Abstract
Version documents are growing faster nowadays. Answering time-travel phrase queries is getting more and more necessary. There are lots of methods to improve the performance of answering time-travel phrase queries. In this paper, we reference the implementation of "Space-Efficient Index Framework for Time-Travel Phrase Query on Versioned Document". However, we find that Kuo's implementation use lots of index space, so we want to reduce the index space. In order to achieve this purpose, we have two implementations, Impl1: compressed suffix tree + treating versions as different documents and Impl2: compressed suffix tree + K^2-Treaps. Comparing with the total index of Kuo's implementation, Imple1 reduced by 63% and Impl2 reduced by 80%, and these two implementations also support time-travel phrase queries and Top-k queries.