Annotations

Annotations provide information about a span text.

Annotations have a label and have features providing information about the text.

For example, to understand the text, the grammatical building blocks such as parts of speech, phrase chunks and sentences can be identified, annotated and examined to determine meaning. Text references in Documents is a type of annotation.

The beginning and end of an annotation is identified using the Text Graph.

See Text Graph, Tokens and Nodes.

Annotations are created during entity extraction. There are annotations that are created by Sintelix by default.

Dictionaries and Entity Extraction Scripts (EES) create annotations. EES can create connections (links) between annotations. EES can use annotations created by Sintelix, Dictionaries and other EES.

Annotations are grouped and given a label, referred to as a namespace.

In this example, the namespaces shown are default, chunk, lookup and pos.

The chunk annotation is only created when the option is enabled in the Document Processing configuration.

When you add additional annotations during entity extraction, when Dictionaries and Entity Extraction Scripts (EES) are applied, more namespaces can be created. Dictionaries and EES can assign custom namespace labels.

Annotations with the namespace label "Tag" will be displayed in Documents as Text References, highlighted with a colour.

The Text Graph Analyser topic provides a video to demonstrate the behaviour of the Tag namespace and other annotations.

The Text Graph Analyser allows you to visualise and explore the Text Graph and Annotations applied to text.

The diagram below shows a sentence displayed in the Text Graph Analyser, which shows a breakdown of the Text Graph and the annotations applied to the sentence. The Text Graph Analyser is an powerful tool that is used to edit and test Dictionaries and EES.

The table below summarises the annotation namespaces commonly created during entity extraction.

The example below illustrates the default language building blocks annotated on the text graph:

Text Graph Analyser: Default Annotations

The diagram shows the default annotations in the Text Graph Analyser.

The example below illustrates the parts of speech (pos) annotated on the text graph:

Text Graph Analyser: pos Annotations

The example below shows the parts of speech (pos) annotations in the Text Graph Analyser.

Revision: Parts of Speech

Below is a quick revision of the primary parts of speech, as a little revision.

There are nine principle parts of speech:

Part	Description	Examples
Nouns	Name of a place, thing or idea	lady, road, tree, love, dream, pencil, Bob
Pronouns	Take the place of a noun	he, she, it, you, we
Verbs	Express action or a state of being	ran, flew, trips, was, is, showed, jumps
Adjectives	Describe a noun	new, shiny, fat, thorough, funny
Adverbs	Describe or modify a verb, adjective, or another adverb	quickly, repeatedly, precisely (often end in -ly)
Prepositions	Establish a relationship between the noun and the rest of the sentence	over, around, through, during, according to, above, beside
Conjunctions	Connect words, phrases and clauses	and, but, or, nor, however, moreover
Interjections	Express surprise or other emotions	Ouch! Eek! Ow! (often have an exclamation point)
Articles	Help define nouns	the, an, a

pos codes

The table below lists the parts of speech (pos) codes.

Abbreviation	Meaning	Example
CC	coordinating conjunction	and
CD	cardinal digit	1, third
DT	determiner	the
EX	existential there	there is
FW	foreign word	les
IN	preposition/subordinating conjunction	in, of, like
JJ	adjective	green large
JJR	adjective, comparative	greener larger
JJS	adjective, superlative	greenest largest
LS	list marker	1)
MD	modal	could, will
NN	noun, singular	cat, tree
NNS	noun plural	cats, trees
NNP	proper noun, singular	American
NNPS	proper noun, plural	Americans
PDT	predeterminer	all, both, half
POS	possessive ending	parent's friend's
PRP	personal pronoun	I, he, it
PRP$	possessive pronoun	my, his, our
PUNC	Punctuation marks	. , ! ?
RB	adverb	little, quietly, however, usually, hard
RBR	adverb, comparative	less, more quietly, harder
RBS	adverb, superlative	least, most quietly, hardest
RP	particle	about, across, through, in, out, over
SYM	Symbol	/ { = *
TO	infinite marker	to
UH	interjection	goodbye
VB	verb	ask
VBD	verb past tense	was, were, pleaded
VBG	verb gerund	being, pleading
VBN	verb past participle	been, reunified
VBP	verb, present tense not 3rd person singular	am, are
VBZ	verb, present tense with 3rd person singular	is
WDT	wh-determiner	which
WP	wh- pronoun (who)	who, what
WP$	Possessive wh-pronoun	whose, whosever
WRB	wh- adverb	when, where, why

The chunk annotation identifies the phrases within a sentence, which are derived from the parts of speech. The example below illustrates the both the parts of speech and chunks annotated on the text graph:

Chunk annotations are only created if the Document Processing configuration has the Enable Chunker checkbox selected. See Phrase Chunker.

Text Graph Analyser: Chunk Annotations

The example below shows the Chunk (phrase) annotations in the Text Graph Analyser.

chunk codes

The table below summarises the chunk codes.

Annotation	Description
SBAR	Clause introduced by a (possibly empty) subordinating conjunction.
ADJO	Adjective Phrase.
ADVP	Adverb Phrase.
CONJP	Conjunction Phrase.
INTJ	Interjection. Corresponds approximately to the part-of-speech tag UH.
LST	List marker. Includes surrounding punctuation.
NP	Noun Phrase.
PP	Prepositional Phrase.
PRT	Particle. Category for words that should be tagged RP.
UCP	Unlike Coordinated Phrase.
VP	Verb Phrase.