-
Notifications
You must be signed in to change notification settings - Fork 2
/
config.py
90 lines (76 loc) · 3.01 KB
/
config.py
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
85
86
87
88
89
90
from itertools import product
from inspirehep.utils.record import get_value
def get_exact_queries(inspire_record):
dois = get_value(inspire_record, 'dois.value')
arxiv_eprints = get_value(inspire_record, 'arxiv_eprints.value')
report_numbers = get_value(inspire_record, 'report_numbers.value')
return [
{'type': 'exact', 'match': 'dois.value.raw', 'values': dois},
{'type': 'exact', 'match': 'arxiv_eprints.value.raw', 'values': arxiv_eprints},
{'type': 'exact', 'match': 'report_numbers.value.raw', 'values': report_numbers}
]
def get_fuzzy_queries(inspire_record):
mini_record = get_mlt_record(inspire_record)
return [{'type': 'fuzzy', 'match': mini_record}]
def get_mlt_record(inspire_record):
"""Returns a small record to be used with ElasticSearch
More Like This query."""
records = []
if inspire_record.get('titles'):
records.append(
{
'titles': inspire_record['titles'],
'boost': 20
}
)
if inspire_record.get('abstracts'):
records.append(
{
'abstracts': inspire_record['abstracts'],
'boost': 20
}
)
if inspire_record.get('report_numbers'):
records.append(
{
'report_numbers': inspire_record['report_numbers'],
'boost': 10
}
)
if inspire_record.get('authors'):
records.append(
{
'authors': inspire_record['authors'][:3]
}
)
return records
def validator(record, result):
"""Validate results to avoid false positives."""
from inspire_json_merger.comparators import AuthorComparator
author_score = 0.5
if record.get('authors') and result.record.get('authors'):
try:
authors_record = min(len(record['authors']), 5)
authors_match_record = min(len(result.record['authors']), 5)
matches = len(
AuthorComparator(
record['authors'][:authors_record],
result.record['authors'][:authors_match_record]
).matches
)
author_score = matches / float(max(authors_record, authors_match_record))
except:
# FIXME json_merger fails internally in some author comparison
pass
title_max_score = 0.5
if record.get('titles') and result.record.get('titles'):
record_titles = [r['title'].lower() for r in record['titles']]
result_titles = [r['title'].lower() for r in result.record['titles']]
for titles in product(record_titles, result_titles):
record_tokens = set(titles[0].split())
result_tokens = set(titles[1].split())
title_score = len(record_tokens & result_tokens) / \
float(len(record_tokens | result_tokens))
if title_score > title_max_score:
title_max_score = title_score
return (author_score + title_max_score) / 2 > 0.5