en-es
en
en-fr
en-sl
0.25
0.5
0.75
1.25
1.5
1.75
2
Querying Wikidata: Comparing SPARQL, Relational and Graph Databases
Published on 2016-11-101364 Views
In this paper, we experimentally compare the efficiency of various database engines for the purposes of querying the Wikidata knowledge-base, which can be conceptualised as a directed edge-labelled gr
Related categories
Presentation
Querying Wikidata: Comparing SPARQL, Relational and Graph Databases00:00
What is Wikidata?00:06
Problems with Wikipedia00:33
Problem 100:37
Problem 200:56
Solution: Wikidata01:07
Wikidata expansion in four years01:27
Wikidata Query Service (WDQS01:37
Example query - 101:43
Example query - 201:53
Results02:05
Thousands queries per minutes02:10
Our main question02:21
A Wikidata Example03:05
Alexis Sanchez03:09
Models, Languages, Engines - 103:50
Models, Languages, Engines - 203:52
Modeling Wikidata as RDF04:54
Objects, predicates and qualifiers - 104:56
Objects, predicates and qualifiers - 205:04
Objects, predicates and qualifiers - 305:12
Standard Reification (SR)05:18
N-ary relations (NR)05:44
Singleton Properties (SP)06:21
Named Graphs (NG)06:38
Modeling Wikidata as Property Graphs06:50
Property Graph06:55
Reified property graph07:53
Modeling Wikidata as a Relational Database08:22
Relational model (1/3)08:23
Relational model (2/3)09:31
Relational model (3/3)09:43
Experiment 1: Atomic Lookups10:02
Statement pattern10:05
Atomic lookups10:47
High selectivity - 111:26
High selectivity - 212:43
High selectivity - 313:08
Low selectivity13:43
Average14:11
Experiment 2: Snowflake / Star Queries14:13
Snowflake queries depth 114:18
Snowflake queries depth 214:51
Results Snowflake queries14:58
Conclusions15:14
Other consideration16:12