Integrated development tool for building a natural language understanding application
First Claim
Patent Images
1. A method of developing natural language understanding (NLU) applications comprising:
- determining NLU interpretation information from an NLU training corpus of text using a multi-pass processing technique, wherein alteration of one pass automatically alters an input for a subsequent pass, said NLU interpretation information specifying an interpretation of at least part of said NLU training corpus of text;
storing said NLU interpretation information in a database;
in a graphical editor, presenting selected items of NLU interpretation information as a meaning tree comprising terminal and non-terminal nodes representative of data items and receiving user specified edits to said NLU interpretation information;
determining an intersection of selected nodes of said meaning tree from said NLU interpretation information;
presenting said intersection of selected nodes as choices for adding an additional node to said meaning tree; and
adding a node above said selected nodes of said meaning tree, wherein said added node is selected from said choices.
1 Assignment
0 Petitions
Accused Products
Abstract
A method of developing natural language understanding (NLU) applications can include determining NLU interpretation information from an NLU training corpus of text using a multi-pass processing technique. The alteration of one pass automatically can alter an input for a subsequent pass. The NLU interpretation information can specify an interpretation of at least part of the NLU training corpus of text. The NLU interpretation information can be stored in a database, and selected items of the NLU interpretation information can be presented in a graphical editor. User specified edits also can be received in the graphical editor.
57 Citations
73 Claims
-
1. A method of developing natural language understanding (NLU) applications comprising:
-
determining NLU interpretation information from an NLU training corpus of text using a multi-pass processing technique, wherein alteration of one pass automatically alters an input for a subsequent pass, said NLU interpretation information specifying an interpretation of at least part of said NLU training corpus of text; storing said NLU interpretation information in a database; in a graphical editor, presenting selected items of NLU interpretation information as a meaning tree comprising terminal and non-terminal nodes representative of data items and receiving user specified edits to said NLU interpretation information; determining an intersection of selected nodes of said meaning tree from said NLU interpretation information; presenting said intersection of selected nodes as choices for adding an additional node to said meaning tree; and adding a node above said selected nodes of said meaning tree, wherein said added node is selected from said choices. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 25, 26, 27, 28, 29, 30, 31, 32, 33)
-
-
24. The method of 23, further comprising:
if a score of said correct interpretation is greater than a score of said predicted interpretation, indicating that said statistical model is incorrect.
-
34. A method of developing natural language understanding (NLU) applications comprising:
-
determining NLU interpretation information from an NLU training corpus of text using a multi-pass processing technique, wherein alteration of one pass automatically alters an input for a subsequent pass, said NLU interpretation information specifying an interpretation of at least part of said NLU training corpus of text; storing said NLU interpretation information in a database; in a graphical editor, presenting selected items of NLU interpretation information as a meaning tree comprised of terminal and non-terminal nodes representing data items and receiving user specified edits to said NLU interpretation information; identifying an intersection of data items; and presenting said identified data items as selections for annotating a user specified node of said meaning tree for a sentence of said NLU training corpus of text.
-
-
35. An integrated development tool for developing a natural language understanding (NLU) application, said integrated development tool comprising:
-
a database configured to store items of NLU interpretation information corresponding to an NLU training corpus of text, wherein said items of NLU interpretation information specify an interpretation of at least part of said NLU training corpus of text; a graphical editor having a plurality of views for manipulating said items of NLU interpretation information, said graphical editor being communicatively linked to said database, wherein a first one of said plurality of views comprises selected items of NLU interpretation information presented as a meaning tree comprised of terminal and non-terminal nodes representing data items, wherein a second one of said plurality of views comprises presenting an intersection of selected nodes of said meaning tree as choices for adding an additional node to said meaning tree; and a processor configured to determine said items of NLU interpretation information according to a multi-pass system, and to store said items of NLU interpretation information in said database, wherein alteration of one pass automatically alters an input for a subsequent pass, wherein said processor being further configured to determine the intersection of selected nodes of said meaning tree from said items of NLU interpretation information, and to add a node above said selected nodes to said meaning tree, wherein said added nodes are selected from said choices. - View Dependent Claims (36, 37, 38, 39)
-
-
40. A machine-readable storage, having stored thereon a computer program having a plurality of code sections executable by a machine for causing the machine to perform the steps of:
-
determining NLU interpretation information from an NLU training corpus of text using a multi-pass processing technique, wherein alteration of one pass automatically alters an input for a subsequent pass, said NLU interpretation information specifying an interpretation of at least part of said NLU training corpus of text; storing said NLU interpretation information in a database; in a graphical editor, presenting selected items of NLU interpretation information as a meaning tree comprised of terminal and non-terminal nodes representing data items and receiving user specified edits to said NLU interpretation information; determining an intersection of selected nodes of said meaning tree from said NLU interpretation information; presenting said intersection of selected nodes as choices for adding an additional node to said meaning tree; and adding a node above said selected nodes of said meaning tree, wherein said added node is selected from said choices. - View Dependent Claims (41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 64, 65, 66, 67, 68, 69, 70, 71, 72)
-
-
63. The machine-readable storage of 62, further comprising:
if a score of said correct interpretation is greater than a score of said predicted interpretation, indicating that said statistical model is incorrect.
-
73. A machine-readable storage having stored thereon a computer program having a plurality of code sections executable by a machine for causing the machine to perform the steps of:
-
determining NLU interpretation information from an NLU training corpus of text using a multi-pass processing technique, wherein alteration of one pass automatically alters an input for a subsequent pass, said NLU interpretation information specifying an interpretation of at least part of said NLU training corpus of text; storing said NLU interpretation information in a database; in a graphical editor, presenting selected items of NLU interpretation information as a meaning tree comprised of terminal and non-terminal nodes representing data items and receiving user specified edits to said NLU interpretation information; identifying an intersection of data items; and presenting said identified data items as selections for annotating a user specified node of said meaning tree for a sentence of said NLU training corpus of text.
-
Specification