Presentation is loading. Please wait.

Presentation is loading. Please wait.

A Fast Regular Expression Indexing Engine Junghoo “John” Cho (UCLA) Sridhar Rajagopalan (IBM)

Similar presentations


Presentation on theme: "A Fast Regular Expression Indexing Engine Junghoo “John” Cho (UCLA) Sridhar Rajagopalan (IBM)"— Presentation transcript:

1 A Fast Regular Expression Indexing Engine Junghoo “John” Cho (UCLA) Sridhar Rajagopalan (IBM)

2 Junghoo "John" Cho (UCLA Computer Science)2 Problem How can we match a regular expression fast? Large text-corpus Several days to match a simple regular expression! Our solution Use an index!

3 Junghoo "John" Cho (UCLA Computer Science)3 Motivation Advanced search interface What is the middle name of Thomas Edison? State-of-the-art: Keyword-based  Thomas Edison Regular expression  Thomas [a-z]+ Edison Data extraction [Brin 98]

4 Junghoo "John" Cho (UCLA Computer Science)4 Outline Index key selection Useful gram Algorithm for key selection Other issues Experiments

5 Junghoo "John" Cho (UCLA Computer Science)5 Motivating example All mp3 URLs on the Web: Every matching string contains mp3. Questions: Should we index “ mp3 ”? Should we index “ <a href= ”?

6 Junghoo "John" Cho (UCLA Computer Science)6 What index entires? Solution 1: Inverted index (English words) Cannot handle many regular expressions Solution 2: k-grams for k = 1, 2, …, 10 Index too large (10 times as large!) Our solution: multigram

7 Junghoo "John" Cho (UCLA Computer Science)7 Main idea “ mp3 ” is helpful. Not many pages have it. “ <a href= ” is not. All pages have it. We index only “useful” grams.

8 Junghoo "John" Cho (UCLA Computer Science)8 Gram selectivity Sel(x): selectivity of gram x Sel(x) = M(x)/N M(x): number of pages containing gram x N: total number of pages C-useful gram: All grams with Sel(x) < C C: system parameter random access vs. sequential access time We index only “C-useful” grams

9 Junghoo "John" Cho (UCLA Computer Science)9 Minimal useful gram “ Unix is great ” If “ Unix ” is useful “ Unix i ”, “ Unix is ”, “ Unix is g ”, … are all useful. “ Unix ” is the minimal useful gram. We index only the minimal useful gram.

10 Junghoo "John" Cho (UCLA Computer Science)10 Advantages Versatile We can look up “ Unix ” for all grams like “ Unix i ”, “ Unix is g ”, etc. Easy to find Reduction to “A priori” algorithm Index size guarantee

11 Junghoo "John" Cho (UCLA Computer Science)11 Algorithm Main idea: If “ abcde ” is minimal useful gram, then “ abcd ” is not useful. If “ abcd ” is not useful, then “ a ”, “ ab ”, “ abc ” is not useful. Minimal useful gram identification is equivalent to useless gram identification.

12 Junghoo "John" Cho (UCLA Computer Science)12 A priori algorithm Useless gram identification Find all sequences of characters that occur in more than k pages A priori algorithm Find all sets of items that occur in more than k baskets Less than 4 scans of the corpus to find all minimal useful grams.

13 Junghoo "John" Cho (UCLA Computer Science)13 Prefix free set A set of grams X is prefix free if no x  X is a prefix of any other x’  X e.g.) X = {ab, ac, abc} is not prefix free. A set of minimal useful grams is a prefix free set.

14 Junghoo "John" Cho (UCLA Computer Science)14 Size of a prefix free set Let X be a set of grams extracted from corpus D and is prefix free. Then |X|  |D| |X|: number of grams in X |D|: number of characters in D The size of an index with minimal useful grams does not exceed the size of the corpus!

15 Junghoo "John" Cho (UCLA Computer Science)15 Shortest suffix gram <a href=“k If =“k is useful, then <a href=“k, a href=“k, href=“k,etc are all useful. =“k: shortest suffix gram We index only the shortest suffix gram. Pre-suf shell

16 Junghoo "John" Cho (UCLA Computer Science)16 Other issues Given a regular expression how to find an index entry to look up? Optimization?

17 Junghoo "John" Cho (UCLA Computer Science)17 Experiments Half million Web documents Comparison Raw scanning Multigram index Complete: k-grams for k = 1,2, …, 10 Benchmark queries No standard Collected from IBM Almaden researchers

18 Junghoo "John" Cho (UCLA Computer Science)18 Example queries (simplified) MP3 URLs: Invalid HTML: ]*< Phone numbers: (\d\d\d) \d\d\d-\d\d\d\d PowerPC chip number: (xpc|mpc)[0-9]+[0-9a-z]+ Middle name of Clinton: William [a-z]+ Clinton

19 Junghoo "John" Cho (UCLA Computer Science)19 Evaluation metrics Index construction time Index size Matching time Overall throughput Response time for first 10 matches

20 Junghoo "John" Cho (UCLA Computer Science)20 Construction time & Index size Complete Multigram Construction Time63 hours6 hours No of Keys103,151,30264,656 No of Postings18,193,048,399820,396,717 An order of magnitude reduction in index size

21 Junghoo "John" Cho (UCLA Computer Science)21 Matching time On average, Complete is faster than Multigram only by 33% Query Scanning Complete Multigram mp3573 sec11 sec15 sec PowerPC548 sec1 sec2 sec phone540 sec

22 Junghoo "John" Cho (UCLA Computer Science)22 Result size & Improvement

23 Junghoo "John" Cho (UCLA Computer Science)23 Related work Suffix tree Beaza-Yates et al., JACM,1998 Main-memory based Disk-based string index Cooper et al., VLDB, 2001 Good for exact string matching Inverted index English words

24 Junghoo "John" Cho (UCLA Computer Science)24 Conclusion Fast matching of regular expressions Multigram index Small size Significant improvement in matching time Future work Optimization?


Download ppt "A Fast Regular Expression Indexing Engine Junghoo “John” Cho (UCLA) Sridhar Rajagopalan (IBM)"

Similar presentations


Ads by Google