Επεξεργασία χωροκειμενικών ερωτημάτων εύρους για μεγάλα δεδομένα
View/ Open
Keywords
Χωρο-κειμενικά δεδομένα ; Χωρο-κειμενική ευρετηρίαση ; Ερωτήματα εύρους ; Κατανεμημένη επεξεργασίαAbstract
The widespread use of the Internet via GPS enabled mobile devices has led to the
generation of an enormous volume of spatio-textual data, characterized by geographic
location and textual description. The complexity of managing spatio-textual data arises
from the high dimensionality of the representation space.
Given the vast volume of spatio-textual data, a variety of query types have been proposed
to address diverse needs. This work, however, concentrates on spatio-textual range
queries. The purpose of such a query is to identify objects that are near to a specified
location and exhibit similarity with the query's keywords. Furthermore, due to the high
volume of data, distributed systems become necessary for effective processing and
storage.
This dissertation presents a spatio-textual indexing algorithm for conducting spatio-
textual range queries within a distributed processing environment. The algorithm was
implemented on Apache Spark and compared with spatial first and textual first methods
in Apache Spark, Apache Sedona and GeoMesa in terms of execution time.