Review of report LSDX A New Labeling Scheme

Review of report "LSDX: A New Labeling Scheme for Dynamically Updating XML Data" 1

Introduction n Any Problem there? n Here is A Solution! 2

Motivation n Several path indexing, labeling and number scheme have been introduced to facilitate query processing for XML data Most of these approaches need to recompute existing labels – time consuming Want to implement a new Labeling Scheme for Dynamic XML data (LSDX) that u u Supports updating XML data without relabeling existing labels Supports the representation of relationships between nodes 3

Aim n - Aim: To provide fast query processing for dynamic XML data Fast update Fast searching Fast insertion and deletion 4

Aim Cont n - - To meet the aim, a new labeling scheme is proposed to: Eliminate overhead of re-labeling existing nodes -> fast insertion and update Fast search and response to structural queries, e. g. relationship between nodes, by having more information on labels 5

Key Findings n Improvement in performance 1. 2. 3. n Relationship suggestion u n Length of labels Time used to generate labels Time used in insertion and deletion ability to show different relationships Tree depth indication u Indication of tree depth of nodes 6

Prove in Theory n Algorithm for indexing u The red-black tree (a. k. a. symmetric binary B-tree ) is employed, which shall guarantee that basic operations take log(n) time in the worst case 7

Prove in Practice 1 n Experiments 1. 2. Length of Labels: much shorter than The two previous major schemas (GRP, SP) Time used to generate labels: one second for 1. 2 MB, one minute for 100 MB 8

Prove in Practice 2 3. Insertion and Deletion Time: Updating XML date is relatively fast without the effort of re-labeling 9

Summary n LSDX is the new labeling scheme introduced in the paper u n 3 main features Review We have pointed out some basic information about LSDX u Found the key points of the report u 10