US2016328381A1PendingUtilityA1

Method and apparatus for referring expression generation

51
Assignee: ARRIA DATA2TEXT LTDPriority: Aug 30, 2012Filed: Apr 7, 2016Published: Nov 10, 2016
Est. expiryAug 30, 2032(~6.1 yrs left)· nominal 20-yr term from priority
G06F 40/56G06F 40/295G06F 40/289G06F 17/2775G06F 17/278
51
PatentIndex Score
0
Cited by
0
References
0
Claims

Abstract

Methods, apparatuses, and computer program products are described herein that are configured to perform referring expression generation. In some example embodiments, a method is provided that comprises identifying an intended referent to be referred to in a textual output. The method of this embodiment may also include determining that a salient ancestor of the intended referent is lower in a part-of hierarchy than a lowest common ancestor. The method of this embodiment may also include causing the salient ancestor to be set as a current target referent and a new salient ancestor to be determined for the current target referent. In some example embodiments, the default descriptor of each current target referent is added to the referring noun phrase and the part-of hierarchy is traversed via salient ancestor links until the new salient ancestor of the current target referent is higher than or equal to the lowest common ancestor.

Claims

exact text as granted — not AI-modified
1 - 20 . (canceled) 
     
     
         21 . A natural language generation method for generating a referring noun phrase for an intended referent found in one or more messages, the method comprising:
 arranging, using a processor, one or more messages, wherein messages represent a phrase or a simple sentence and are created in an instance in which an input data stream comprises data that satisfies one or more message requirements, and wherein at least a portion of the input data stream comprises non-linguistic data;   determining a lowest common ancestor for an intended referent in a message of the one or more messages;   generating a referring noun phrase for the intended referent to be included in a textual output by traversing a part-of hierarchy;   generating the textual output comprising the referring noun phrase such that it is displayable on a user interface, wherein the textual output linguistically describes at least a portion of the input data stream; and   displaying the textual output via a display device.   
     
     
         22 . A method according to  claim 21 , wherein the one or more parts of the part-of hierarchy that are traversed based on one or more salient ancestor links. 
     
     
         23 . A method according to  claim 21 , further comprising:
 determining that the intended referent is marked as salient; and   causing the referring noun phrase to solely comprise the default descriptor of the intended referent.   
     
     
         24 . A method according to  claim 21 , further comprising:
 determining that the salient ancestor is equal to the lowest common ancestor; and   causing the referring noun phrase to comprise the default descriptor of the intended referent.   
     
     
         25 . A method according to  claim 21 , further comprising:
 determining the previously referred-to entity based on a last mentioned entity in a discourse model.   
     
     
         26 . A method according to  claim 25 , wherein the previously referred-to entity is set to a root component of the part-of hierarchy in an instance in which the previously referred-to entity is set to null. 
     
     
         27 . A method according to  claim 21 , wherein the default descriptor of an entity is at least one of a default descriptor, a class name or a type name. 
     
     
         28 . A method according to  claim 21 , further comprising:
 determining that one or more parts of the part-of hierarchy traversed are marked as ignore, wherein a default descriptor of the one or more parts of the part-of hierarchy that are marked as to be ignored are not included in the referring noun phrase.   
     
     
         29 . A method according to  claim 21 , wherein the default descriptor of an entity further comprises at least one of a class name or a type name. 
     
     
         30 . A method according to  claim 21 , wherein the referring noun phrase comprises a predetermined maximum number of premodifiers of the default descriptor of the intended referent, wherein the predetermined number of premodifiers comprise one or more default descriptors of the one or more parts of the part-of hierarchy that are traversed. 
     
     
         31 . A method according to  claim 30 , wherein the referring noun phrase comprises a set of postmodifiers of the default descriptor of the intended referent, wherein the set of postmodifiers comprise one or more default descriptors of one or more parts of the part-of hierarchy that are traversed that were not included as premodifiers. 
     
     
         32 . An apparatus comprising:
 at least one processor; and   at least one memory including computer program code, the at least one memory and the computer program code configured to, with the at least one processor, cause the apparatus to at least:   arrange one or more messages, wherein messages represent a phrase or a simple sentence and are created in an instance in which an input data stream comprises data that satisfies one or more message requirements, wherein at least a portion of the input data stream comprises non-linguistic data;   determine a lowest common ancestor for an intended referent in a message of the one or more messages;   generate a referring noun phrase for a salient ancestor of the intended referent to be included in a textual output by traversing a part-of hierarchy;   generate the textual output comprising the referring noun phrase such that it is displayable on a user interface, wherein the textual output linguistically describes at least a portion of the input data stream; and   display the textual output via a display device.   
     
     
         33 . An apparatus according to  claim 32 , wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:
 determine that the intended referent is marked as salient; and   cause the referring noun phrase to solely comprise the default descriptor of the intended referent.   
     
     
         34 . An apparatus according to  claim 32 , wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:
 determine that the salient ancestor is equal to the lowest common ancestor; and   cause the referring noun phrase to comprise the default descriptor of the intended referent.   
     
     
         35 . An apparatus according to  claim 32 , wherein the at least one memory including the computer program code is further configured to, with the at least one processor, cause the apparatus to:
 determine the previously referred-to entity based on a last mentioned entity in a discourse model.   
     
     
         36 . A non-transitory computer readable memory medium having program code instructions stored thereon, the program code instructions which when executed by an apparatus cause the apparatus at least to:
 arrange one or more messages, wherein messages represent a phrase or a simple sentence and are created in an instance in which an input data stream comprises data that satisfies one or more message requirements, wherein at least a portion of the input data stream comprises non-linguistic data;   determine a lowest common ancestor for an intended referent in a message of the one or more messages;   generate a referring noun phrase for the intended referent to be included in a textual output by traversing a part-of hierarchy;   generate the textual output comprising the referring noun phrase such that it is displayable on a user interface, wherein the textual output linguistically describes at least a portion of the input data stream; and   display the textual output via a display device.   
     
     
         37 . A computer program product according to  claim 36 , further comprising program code instructions configured to:
 determine that the intended referent is marked as salient; and   cause the referring noun phrase to solely comprise the default descriptor of the intended referent.   
     
     
         38 . A computer program product according to  claim 36 , further comprising program code instructions configured to:
 determine that the salient ancestor is equal to the lowest common ancestor; and   cause the referring noun phrase to comprise the default descriptor of the intended referent.   
     
     
         39 . A computer program product according to  claim 36 , further comprising program code instructions configured to:
 determine the previously referred-to entity based on a last mentioned entity in a discourse model.   
     
     
         40 . A computer program product according to  claim 39 , wherein the previously referred-to entity is set to a root component of the part-of hierarchy in an instance in which the previously referred-to entity is set to null. 
     
     
         41 . A method according to  claim 21 , wherein generating the referring noun phrase further comprises:
 removing a first element in the queue, wherein the first element is designated as a head noun in the referring noun phase.   
     
     
         42 . A method according to  claim 41 , wherein generating the referring noun phrase further comprises:
 removing a next element in the queue; and   setting the next element as a premodifier to the head noun; and   in instance in which a predetermined premodifier count threshold is satisfied, removing an element from the queue and setting it as a post modifier to the head noun.

Cited by (0)

No later patents cite this yet.

References (0)

No backward citations on record.