verbnetframe

class verbnetframe.ComputeSlotTypeMixin[source]

Bases: builtins.object

_abc_cache = <_weakrefset.WeakSet object at 0x7fd4f39daf98>
_abc_negative_cache = <_weakrefset.WeakSet object at 0x7fd4f39daef0>
_abc_negative_cache_version = 25
_abc_registry = <_weakrefset.WeakSet object at 0x7fd4f39dac50>
static _is_a_slot(elem)[source]

Tell wether an element represent a slot

Parameters:elem (str.) – The element.
Returns:bool – True if elem represents a slot, False otherwise
compute_slot_types(structure)[source]

Build the list of slot types for this frame

slot_types = {'indirect_object': 'OBJI', 'subject': 'SBJ', 'prep_object': 'PPOBJ', 'object': 'OBJ'}
class verbnetframe.VerbnetFrameOccurrence(structure, roles, predicate)[source]

Bases: verbnetframe.ComputeSlotTypeMixin

A representation of a FrameNet frame occurrence converted to VerbNet representation for easy comparison.

Variables:
  • structure – (str | str set) list – representation of the structure
  • roles – set list – possible VerbNet roles for each structure’s slot
  • num_slots – int – number of argument slots in :structure
  • predicate – str – the predicate
  • headwords – str – the head word of each argument
_abc_cache = <_weakrefset.WeakSet object at 0x7fd4f39e9b38>
_abc_negative_cache = <_weakrefset.WeakSet object at 0x7fd4f39e9b70>
_abc_negative_cache_version = 25
_abc_registry = <_weakrefset.WeakSet object at 0x7fd4f39e9b00>
static _keep_only_keywords(sentence)[source]

Keep only keywords and phrase type markers in the structure

Parameters:sentence (str.) – The structure to reduce.
Returns:String – the reduced string
static _reduce_args(frame, structure, new_begin)[source]

Replace the predicate and the argument of a frame by phrase type marks

Parameters:
  • frame (Frame.) – The original Frame.
  • structure (str.) – The current structure representation.
  • new_begin (int.) – The left offset cause by previous manipulations.
Returns:

String – the reduced string
static build_from_frame(gold_framenet_instance, conll_frame_instance)[source]

Build a VerbNet frame from a FrameInstance object

Parameters:
  • gold_framenet_instance – The gold FrameNet frame instance
  • conll_frame_instance – The frame instance from CoNLL
Returns:

VerbnetFrameOccurrence – the frame without the gold roles

converted to VerbNet-style representation

phrase_replacements = {'QUO': 'S', 'N': 'NP', 'Sinterrog': 'S', 'Poss': 'NP', 'VPto': 'to S', 'VPing': 'S_ING', 'VPbrst': 'S', 'Sfin': 'S'}
class verbnetframe.VerbnetOfficialFrame(structure, roles, vnclass, role_restrictions)[source]

Bases: verbnetframe.ComputeSlotTypeMixin

A representation of a frame syntactic structure

Variables:
  • structure – (str | str set) List – representation of the structure
  • roles – str list – VerbNet roles for each structure’s slot
  • num_slots – int – number of argument slots in :structure
  • vnclass – str – the class number, eg. 9.10
  • example – str – An example sentence that illustrates the frame
_abc_cache = <_weakrefset.WeakSet object at 0x7fd4f39e9ba8>
_abc_negative_cache = <_weakrefset.WeakSet object at 0x7fd4f39e9c18>
_abc_negative_cache_version = 25
_abc_registry = <_weakrefset.WeakSet object at 0x7fd4f39e9be0>
generate_relatives()[source]
passivize()[source]

Based on current frame, return a list of possible passivizations