1. EASE: An Effective 3-in-1 Keyword Search Method for Unstructured, Semi-structured and Structured Data Cuoliang Li, Beng Chin Ooi, Jianhua Feng, Jianyong Wang, Lizhu Zhou Tsinghua University SIGMOD 2008 2009. 03. 19. Summarized by Jaehui Park, IDS Lab., Seoul National University Presented by Jaehui Park, IDS Lab., Seoul National University

2. Copyright  2008 by CEBT INTRODUCTION  Keyword search capability into text documents, XML documents, and relational databases  Graph index Instead of traditional inverted index – Effective for unstructured data – Inadequate for complex structural information.  EASE (Efficient and Adaptive keyword Search method) Efficient algorithmic basis for scalable top-k-style processing of large amounts of heterogeneous data – Employing and adaptive, efficient and novel index 2

3. Copyright  2008 by CEBT Contributions  Model for unstructured, semi-structured and structured data as graphs  Effective graph index as opposed to the inverted index  Novel ranking mechanism for both DB and IR viewpoint  Extensive performance study 3

4. Copyright  2008 by CEBT Motivation  Unstructured Link awareness – Relevant data may be separated into different pages but linked through hyperlinks  (Semi-) Structured LCA (Lowest common ancestors) – Connected tree with minimal cost Ex) Steiner trees 4

5. Copyright  2008 by CEBT r-Radius Steiner Graph Problem  Meaningful Steiner graphs with acceptable sizes  Several concepts Centric distance Radius r-Radius Steiner tree – Radius of a Steiner graph cannot be larger than r 5

6. Copyright  2008 by CEBT Example  DBLP example 6

7. Copyright  2008 by CEBT The r-Radius Seiner Graph Problem  Given a graph and an input keyword query K, the r-Radius Seiner Graph Problem is to find all the r-radius Steiner graphs in, which contain all or a portion of the input keywords in K, ranked by relevancy with K. 7

8. Copyright  2008 by CEBT EASE: An adaptive search method  Inverted indices are not effective for discovering the much richer structural relationships existing in databases with complicated structured [10]. Index r-radius Steiner graphs for each combination – Very expensive  Proposed method 1. Discover r-radius graphs (indexing) 2. Extracting r-radius Steiner graphs (on the fly) – By removing non-Steiner nodes 8

9. Copyright  2008 by CEBT EASE: An adaptive search method  Adjacency Matrix Extracting r-radius graphs effectively 9

10. Copyright  2008 by CEBT EASE: An adaptive search method  Determining the subgraph that are r-radius graphs By Lemma 1. For efficient retrieval of r-radius graphs – Graph index r-radius graph that contain query keywords k  Extracting r-radius Steiner graphs By Theorem 1. 10

11. Copyright  2008 by CEBT EASE: An adaptive search method  Computing the Steiner nodes 11

12. Copyright  2008 by CEBT EASE: An adaptive search method  Maximal r-Radius Graph Avoid redundancy – Keep the maximal r-radius graphs in the graph index Overlapping graphs  Graph partitioning Avoid the incurrence of huge storage Only need to retrieve the corresponding relevant graph partitions Graph similarity – Bigger overlap -> higher similarity 12

13. Copyright  2008 by CEBT Summary  1. Obtain adjacency matrix M  2. Compute M r  3. Extract the maximal r-radius graphs  4. Cluster the graphs by employing the existing K-means algorithm and partition the graph  5. Construct the graph index to materialize the maximal r-radius graphs 13

14. Copyright  2008 by CEBT Others  Ranking Functions TF-IDF based IR-ranking Structural Compactness-based DB Ranking – Intuitively, when an r-radius Steiner graph SG is more compact, SG is more likely to be meaningful and relevant.  Indexing 14

15. Copyright  2008 by CEBT Experimental study  Dataset: DBLife, DBLP and IMDB  Comparison Unstructured – InfoUnit [18] Semi-structured – SLCA [28] Structured – DPBF [6] 15

16. Copyright  2008 by CEBT Experimental study 16

17. Copyright  2008 by CEBT Experimental study 17

18. Copyright  2008 by CEBT Conclusion  Proposed an efficient and adaptive keyword search method EASE – Keyword queries over unstructured, semi-structured and structure data  Examined the issues of indexing and ranking By taking into account both the structural compactness  Experimental results shows that EASE achieves both high search efficiency and quality for keyword search over heterogeneous data. 18