forked from fedormyskin/Linked-Potter
-
Notifications
You must be signed in to change notification settings - Fork 0
/
R-Potter.R
85 lines (72 loc) · 5.88 KB
/
R-Potter.R
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
## This file describes how to retrieve information from the RDF/OWL file containing the knowledge base.
# However, it doesn't work on common computers because the number of entities is too high and requires
# excessive computational power.
## install redland and rdflib
# if you get a configuration error, trying installing redland from a terminal first and then install the R package. For Unix:
# $ brew install pkg-config
# $ brew install redland
# $ export PKG_CONFIG_PATH=$PKG_CONFIG_PATH:~/usr/local/Cellar/redland/1.0.17_1/lib/pkgconfig # (or your local path, if different)
install.packages("redland")
install.packages("rdflib") # excellent introduction to RDF files and use of the package: https://cran.r-project.org/web/packages/rdflib/vignettes/rdf_intro.html
## use RDF/XML file, i.e. the linked data, ontology, knowledge base, or graph database
library(rdflib)
library(tidyverse)
# download the ontology file from the OSF repository: https://osf.io/azu86
# open and parse the RFD/XML file (.rdf, .xml, .owl, .jsonld, .ttl, .obo)
rdf <- rdf_parse("FILEPATH")
## define SPARQL queries
# commented examples of how to use queries available here: https://www.w3.org/TR/sparql11-query/
# note: all terms/individuals are identified by an automatically generated Internationalized Resource Identifier (IRI), so to understand the results you have to print the corresponding rdfs:label value
# basic triples available: CharacterTag participatesTo RelationshipTag, CharacterTag/RelationshipTag isTaggedAs FreeformTag (plus inverse types and subclasses)
#query to get the available classes
query_classes <-
'PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> # prefixes are useful to shorten the query
PREFIX owl: <http://www.w3.org/2002/07/owl#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
SELECT ?subject_label # define which variables/values to print
WHERE {
?subject a owl:Class . # define the type of variable you want to query
?subject rdfs:label ?subject_label # define a variable for the subject label, which will be printed to make the query results human-readable
}'
# query to get the releationships in which a character participates
# modify this to query for other types of triples
# currently there is an issue: if you don't get relationships for a character, query one of the character's synonyms
query_ships <-
'PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> # prefixes are useful to shorten the query
PREFIX owl: <http://www.w3.org/2002/07/owl#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
SELECT ?subject_label ?predicate_label ?object_label # define which variables/values to print
WHERE {
?subject ?predicate ?object. # define an empty triple
?subject rdfs:label ?subject_label . # define a variable for the subject label, which will be printed to make the query results human-readable
?object rdfs:label ?object_label . # define a variable for the object label, which will be printed to make the query results human-readable
?predicate rdfs:label ?predicate_label . # define a variable for the predicate label, which will be printed to make the query results human-readable
?predicate rdfs:label "participatesTo" . # modify this to choose the predicate of the triple you want to query: participatesTo, hasParticipant, isTagOf, isTaggedAs
?isCanonical rdfs:label "isCanonical" . # define a variable for the predicate isCanonical
?subject ?isCanonical true . # optional: show only canonical CharacterTags
# ?object ?isCanonical true . # optional: use this to see only canonical RelationshipTags (data for RelationshipTags with less than 1000 stories are not reliable)
BIND (EXISTS{?subject ?isCanonical ?true} as ?canonical_label) # optional: use this to see if a tag is canonical. Remember to add the variable ?canonical_label in the SELECT line
FILTER regex(str(?subject_label), "Harry Potter") . # modify this to filter the subject you are interested in. Note that all occurrences of the term will be printed, for exact match use WHERE
# FILTER regex(str(?object_label), "Draco Malfoy/Harry Potter") # use this if you need to specify an object of the triple, instead of a subject. Remeber to change the predicate to hasParticipant
}'
# query to get the synonyms of a tag
# modify this to query for other tags
query_synonyms <-
'PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> # prefixes are useful to shorten the query
PREFIX owl: <http://www.w3.org/2002/07/owl#>
PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
PREFIX xsd: <http://www.w3.org/2001/XMLSchema#>
SELECT ?subject_label ?synonym_label #define which variables/values to print
WHERE {
?subject owl:sameAs ?synonym . # define the triple for synonymity
?subject rdfs:label "Harry Potter" . # define the tag for which you want to get the synonyms
?subject rdfs:label ?subject_label . # define a variable for the subject label, which will be printed to make the query results human-readable
?synonym rdfs:label ?synonym_label . # define a variable for the synonym label, which will be printed to make the query results human-readable
#FILTER regex(str(?synonym_label), "fem") . # use and modify this to filter the synonyms you are interested in, e.g. tags for female Harry Potter characters
}'
# perform queries and save outputs
classes <- rdf_query(rdf, query_classes)
HP_ships <- rdf_query(rdf, query_ships)
HP_synonyms <- rdf_query(rdf, query_synonyms)