cosmel.corpus.mention module¶

class cosmel.corpus.mention.Mention(article, sid, mid, *args, gid='', nid='', rid='', rule='', **kwargs)[source]¶

Bases: object

The mention class.

Parameters:

article (Article) – the article containing this mention.
sid (int) – the sentence index in the aritcle.
mid (int) – the mention index in the sentence.
gid (str) – the golden product ID.
nid (str) – the network-predicted product ID.
rid (str) – the rule-labeled product ID.
rule (str) – the rule.
idxs (slice) – the indix slice of this mention.

article¶

the article containing this mention.

Type:	`Article`

sentence¶

the sentence containing this mention.

Type:	`WsWords`

sentence_pre¶

the words before this mention in the sentence.

Type:	`WsWords`

sentence_pre_(with_mention=True)[source]¶: WsWords: the words before this mention in the sentence (with/without mention itself).

sentence_post¶

the words after this mention in the sentence.

Type:	`WsWords`

sentence_post_(with_mention=True)[source]¶: WsWords: the words after this mention in the sentence (with/without mention itself).

mention¶

this mention.

Type:	`WsWords`

bundle¶

the mention bundle containing this mention.

Type:	`MentionBundle`

asmid¶

the tuple of article ID, sentence ID, and mention ID.

Type:	tuple

ids¶

aid¶

the article ID.

Type:	str

sid¶

the sentence ID (the sentence index in the article).

Type:	int

start_idx¶

the starting index of the mention in the sentence.

Type:	int

end_idx¶

the ending index of the mention in the sentence.

Type:	int

last_idx¶

the index of the last word of the mention in the sentence.

Type:	int

mid¶

the mention ID (the mention index in the sentence).

Type:	int

gid¶

the golden product ID.

Type:	str

nid¶

the network-predicted product ID.

Type:	str

rid¶

the rule-labeled product ID.

Type:	str

rule¶

the rule for the product ID.

Type:	str

head_ws¶

the word-segmented head word.

Type:	`WsWords`

head¶

the head word.

Type:	str

head_txt¶

the head word.

Type:	str

head_tag¶

the head post-tag.

Type:	str

head_role¶

the head role.

Type:	str

attrs¶: The xml attributes.

start_xml¶

the starting XML tag.

Type:	str

start_xml_(**kwargs)[source]¶: str: the starting XML tag with custom attributes.

end_xml¶

the ending XML tag.

Type:	str

json¶: Convert to json.

set_gid(gid)[source]¶: Sets the golden product ID.

set_nid(nid)[source]¶: Sets the network-predicted product ID.

set_rid(rid)[source]¶: Sets the rule-labeled product ID.

set_rule(rule)[source]¶: Sets the rule for the product ID.

class cosmel.corpus.mention.MentionSet(mention_bundles)[source]¶

Bases: collections.abc.Collection

The set of mentions.

Item: mention (Mention)

Parameters:	mention_bundles (`MentionBundleSet`) – the set of mention bundles.

path¶

the root path of the mentions.

Type:	str

class cosmel.corpus.mention.MentionBundle(file_path, article)[source]¶

Bases: collections.abc.Sequence

The bundle of mentions in an article.

Item: mention (Mention)

Parameters:	file_path (str) – the path to the mention bundle. article (`Article`) – the article containing this mention bundle.

article¶

the article of this bundle.

Type:	`Article`

aid¶

the article ID (with leading author name and underscore).

Type:	str

path¶

the related file path.

Type:	str

save(file_path)[source]¶: Save the mention bundle to json file.

class cosmel.corpus.mention.MentionBundleSet(mention_root, article_set)[source]¶

Bases: collections.abc.Collection

The set of mention bundles.

Item: mention bundle (MentionBundle)

Parameters:	article_root (str) – the path to the folder containing word segmented article files. mention_root (str) – the path to the folder containing mention files. article_set (`ArticleSet`) – the set of articles.

path¶

the root path of the mentions.

Type:	str

save(output_root)[source]¶: Save all mention bundles to files.

class cosmel.corpus.mention.Id2Mention(mention_set)[source]¶

Bases: collections.abc.Mapping

The dictionary maps article ID, sentence ID, and mention ID to mention object.

Key: the article ID, sentence ID, and mention ID (tuple).
Item: the mention object (Mention).

Parameters:	mention_set (`MentionSet`) – the mention set.

class cosmel.corpus.mention.Id2MentionBundle(id_to_article)[source]¶

Bases: collections.abc.Mapping

The dictionary maps article ID to mention bundle.

Key: the article ID (str).
Item: the mention bundle (MentionBundle).

Parameters:	id_to_article (`Id2Article`) – the dictionary maps article ID to article object.

class cosmel.corpus.mention.Head2MentionList(mention_set)[source]¶

Bases: collections.abc.Mapping

The dictionary maps head word to mention object list.

Key: mention head word (str).
Item: ReadOnlyList of mention object (Mention).

Parameters:	mention_set (`MentionSet`) – the mention set.