Natural language data analytics platform
AssigneeArtificial Solutions Iberia SL
InventorsAndreas Wieweg, David Hjelm, Sonja Petrovic Lundberg, Eric Aili
Life of this patent
Estimating the term…
Abstract
A system for natural language analytics, stored and operating on a network-connected computing device, comprising a natural language application data importer, further comprising a natural language application data importer, a natural language application data augmenter that enriches the data and an analytics component which provides a means of querying structured as well as unstructured data and which also contains a method for providing adaptive natural language analytics.
Claims7
A system for natural language analytics, comprising:
- a natural language application analysis manager comprising at least a processor, a memory, and a plurality of programming instructions stored in the memory and operating on the processor, wherein the programming instructions, when operating on the processor, cause the processor to:
- perform data analysis on the natural language application data using a natural language interaction query language (NLIQL) wherein the data analysis comprises storing natural language application data as sequences of events;
- cross-reference the analyzed data against additional data from external sources;
- model a first natural language session as an ordered sequence of natural language events;
- query the data stored in the system by using NLIQL;
- produce a second natural language session based on the first natural language session and the results of the analysis and query operations;
- provide the second natural language session to an external natural language interaction source for use;
- wherein NLIQL is a query language that provides functionality for:
- analyzing natural language interaction;
- modeling natural language interaction elements as events in a flat file structure, wherein each event belongs to one of three event types, the event types comprising inputs sent to a natural language interaction system, responses sent from a natural language interaction system, and path events comprising actions occurring after an input and prior to a response;
- modeling natural language interaction sessions as ordered sequences of events;
- defining constraints within the unstructured, natural language elements of the data stored in the system;
- defining constraints within the structured, natural language independent elements of the data stored in the system;
- and
- combining these both kinds of constraints in one and same query.
A method for natural language analytics, comprising the steps of:
- (a) receiving, at a natural language application analysis manager comprising at least a processor, a memory, and a plurality of programming instructions stored in the memory and operating on the processor, interaction-specific data from a natural language application;
- (b) performing data analysis on the natural language application data using a natural language interaction query language (NLIQL) wherein the data analysis comprises storing natural language application data as in a flat format as sequences of events;
- (c) cross-referencing the analyzed data against additional data from external sources;
- (d) modeling a first natural language session as an ordered sequence of natural language events;
- (e) querying the data stored in the system by using NLIQL;
- (f) producing a second natural language session based on the first natural language session and the results of the analysis and query operations;
- and
- (g) providing the second natural language session to an external natural language interaction source for use;
- wherein the NLIQL provides functionality for:
- analyzing natural language interaction;
- modeling natural language interaction elements as events in a flat file structure, wherein each event belongs to one of three event types, the event types comprising inputs sent to a natural language interaction system, responses sent from a natural language interaction system, and path events comprising actions occurring after an input and prior to a response;
- modeling natural language interaction sessions as ordered sequences of events;
- defining constraints within the unstructured, natural language elements of the data stored in the system;
- defining constraints within the structured, natural language independent elements of the data stored in the system;
- and
- combining these both kinds of constraints in one and same query.
Dependencies are inferred from the claim text.