Proximity Search in Databases
Roy Goldman, Narayanan Shivakumar,
Suresh Venkatasubramanian, Hector Garcia-Molina
1998
מוצג ע"י: אייל מישור
הקדמה
דוגמה
Internet Movie Database
(www.imdb.com)over 140,000 movies
over 500,000 film industry workers
Queries Structure:
Find <keyword> Near <keyword>
דוגמה - המשך
הבעיה
ranking objects in the Find set based on their proximity to objects in the Near set
ה-
framework של הפתרון
IMDB
IMDB
חישוב ה-
proximityמציאת המרחקים
Self-Joins
Hub Indexing
Constructing Hub Index
בחירת ה-
Hubs
Performance Experiments
S = 10; no more than 2.5% hubs
S = 10; no more than 2.5% hubs
K= 12 no more than 2.5% hubs
S = 10; K = 12
הרחבות