INTELLIGENT MATCHING SYSTEM WITH ONTOLOGY-AIDED RELATION EXTRACTION

• US 20180232443A1
• Filed: 02/16/2017
• Published: 08/16/2018
• Est. Priority Date: 02/16/2017
• Status: Active Application

• Current Assignee: Globality, Inc.
• Original Assignee: Globality, Inc.
• Inventors: Delgo, Lior, Ever-Hadani, Adam, Harpaz, Ran
• Application Number:
  US15/435,070
• Publication Number:
  US 20180232443A1
• Time in Patent Office: Days
• CPC Class Codes: G06F 16/322 Trees
• CPC Class Codes: G06F 16/334 Query execution G06F16/335 takes precedence
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: G06F 16/367 Ontology
• CPC Class Codes: G06F 16/9024 Graphs; Linked lists G06F16/9027 takes precedence
• CPC Class Codes: G06F 40/295 Named entity recognition
• CPC Class Codes: G06Q 30/06 Buying, selling or leasing transactions

Abstract

An intelligent matching system network architecture with a software driven engine may establish one or more matches between properties attributed to an entity, for example, a service provider and a set of specified parameters, for example, the request properties of a customer. The system analyzes user provided free-form text to generate an automated match of requirements specified by a user and a prospective service provider. One or more service providers which are good matches for filling a specified set of parameters, for example, parameters specifying customer requirements or needs derived at least in part from user input are identified.

Claims

1. A method for matching properties to requirements comprising the steps of:

a text-to-ontology mapping input text to an ontology; and

scoring correlations between an output of said mapping step and a plurality of said candidate data sets.

2. A method according to claim 1 further comprising the step of ranking said candidate data sets according to said scoring.

3. A method according to claim 2 further comprising the step of:

defining a knowledge graph and an associated ontology;

generating a taxonomy of named entities correlated to relations of said named entities and associated meta-data;

defining a contextual ontology;

generating a structured representation of parameters of interest in said free-form text input context;

associating a set of language-specific text-matching attributes with nodes in said associated ontology and said contextual ontology and said knowledge graph, in the form of semantically-meaningful dense word embedding vectors;

performing named entity recognition and named entity linking on said input text to associate parts of said input text with named entities of said candidate data sets in said knowledge graph;

analyzing said input text, by performing part-of-speech tagging and dependency parsing on said input text; and

performing a semantic parsing of said input text, extracting semantic relations by traversing a dependency parsed tree and applying a classifier to detect mappings between phrases in said input text and nodes in said contextual ontology.

4. A method according to claim 3 further comprising the step of:

processing named entity to ontology associations to resolve ambiguities and prune matches, to arrive at a context-specific structured representation of said input text.

5. A method according to claim 1 further comprising the step of ranking said candidate data sets according to said scoring.

6. A method according to claim 2 further comprising the steps of:

extract named entities from input text;

ascertain semantic relations between extracted entities and other constituents of the sentence signifying an action (predicate) and a subject;

map said semantic relations to parameters in a contextual ontology.

7. A method according to claim 2 wherein said contextual ontology comprises a graph-based representation of properties of a client entity, a vendor entity, and project requirements.

8. A method according to claim 1 further comprising the step of tokenization of said input text.

9. A method according to claim 8 further comprising the step of part-of-speech analysis and tagging on said input text.

10. A method according to claim 9 wherein said step of part-of-speech analysis and tagging uses a trained conditional random field model.

11. A method according to claim 10 further comprising the step of shallow parsing said input text.

12. A method according to claim 11 further comprising the step of normalizing phrases identified by said step of shallow parsing.

13. A method according to claim 3 further comprising:

assigning one or more word embedding vectors to named entities;

identifying a node in said dependency parsed tree which overlaps a corresponding named entity token in said input text which is the least distance from a root of the parsed tree corresponding to said identified named entity;

traversing the dependency parsed tree upward from said node and establishing a candidate relation predicate by materializing a word embedding vector corresponding to each word encountered along a path to said root; and

determine a minimum distance between any of the embedded word vectors and the word vector associated with the named entity.

14. A method according to claim 14 further comprising the step of backtracking from the root of the tree and traversing down each sub-tree of the dependency parsed tree, in breadth-first search (BFS) order:

and

collecting any word found which is classified as a subject above a certain pre-defined confidence score (probability) threshold.