-
Notifications
You must be signed in to change notification settings - Fork 0
/
__init__.py
109 lines (83 loc) · 2.82 KB
/
__init__.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
91
92
93
94
95
96
97
98
99
100
101
102
103
104
from dataclasses import dataclass
from typing import List, Dict, Optional, Tuple
from numpy import ndarray
@dataclass
class Section:
# dataclass wrapper for section in a document and associated text (split into sentences)
id: str # section name
sentences: List[str]
meta: Optional[Dict] = None
@dataclass
class Document:
# dataclass wrapper for documents yielded by a dataset iterator
sections: List[Section]
reference: List[str]
meta: Optional[Dict] = None
@dataclass
class SentenceEmbeddings:
# dataclass wrapper for section in a document and associated sentence embeddings
id: str # section name
embeddings: ndarray # first dim = number of sentences
meta: Optional[Dict] = None
@dataclass
class SectionEmbedding:
# dataclass wrapper for section in a document and associated embedding
id: str # section name
embedding: ndarray
meta: Optional[Dict] = None
@dataclass
class Embeddings:
# dataclass wrapper for section in a document and associated sentence embeddings
sentence: List[SentenceEmbeddings]
section: List[SectionEmbedding]
meta: Optional[Dict] = None
PairIndices = List[Tuple[int, int]]
@dataclass
class SentenceSimilarities:
# dataclass wrapper for intrasection similarities (sentence to sentence or sentence to section)
id: str # section name
similarities: ndarray
pair_indices: PairIndices
directions: List[str]
meta: Optional[Dict] = None
@dataclass
class SectionSimilarities:
# dataclass wrapper for inter-section similarities (section to section)
similarities: ndarray
pair_indices: PairIndices
directions: List[str]
meta: Optional[Dict] = None
@dataclass
class SectionSentSimilarities:
# dataclass wrapper for inter-section similarities (section to sent)
similarities: ndarray
pair_indices: PairIndices
directions: List[str]
meta: Optional[Dict] = None
@dataclass
class Similarities:
# dataclass wrapper for similarities in a document
sent_to_sent: List[SentenceSimilarities]
sect_to_sect: SectionSimilarities
sent_to_sect: List[SentenceSimilarities]
# sparse_sent_to_sect: List[SentenceSimilarities]
# all_sent_to_sent: SentenceSimilarities
# global_index_mapping: dict
meta: Optional[Dict] = None
@dataclass
class AllSimilarities:
sent_to_sent: List[SentenceSimilarities]
sect_to_sect: SectionSimilarities
sent_to_sect: List[SentenceSimilarities]
# sparse_sent_to_sect: List[SentenceSimilarities]
all_sent_to_sent: SentenceSimilarities
global_index_mapping: dict
meta: Optional[Dict] = None
score = float
section_idx = int
local_idx = int
global_idx = int
sentence = str
Scores = List[Tuple[score,section_idx,local_idx,global_idx]]
Summary = List[Tuple[sentence, score, section_idx, local_idx, global_idx]]
Reference = List[sentence]