Granted Patent B1

US 12,282,897 · App. 18/419,807 · Granted Apr 22, 2025

Inventor: Senthil Manickavasagam Kumar (Dublin, CA)

Assignee: Slate Technologies Inc.

G06Q10/103G06F40/30G06N5/02G06Q10/06312G06Q50/08

Abstract

Claims (68)

1. A method for generating one or more insights related to a building construction project having one or more construction objectives by establishing and generating associative contextual links between data from a plurality of data sources in a computing environment, the method comprising:

obtaining, via a network, the data from the plurality of data sources based on a request received from a client computer for the one or more insights associated with the building construction project, wherein the plurality of data sources corresponds to disparate systems generating unlinked data;

generating a decomposed data set comprising a relevant data set from the obtained data;

parsing and analyzing the decomposed data set based on a set of attribute analyzers to associate one or more attributes to the decomposed data set;

determining an intent of data from the decomposed data set associated with each of the plurality of data sources by applying a natural language parser;

generating a semantic graph of the decomposed data set based on the intent of data to evaluate data relatability between the decomposed data set;

generating atomic knowledge units corresponding to the data based on the parsed decomposed data set and the semantic graph, wherein the atomic knowledge units represent the data in an organized data format;

training one or more machine learning models using training data, wherein the training data includes data relating to the building construction project and one or more user inputs;

analyzing the atomic knowledge units corresponding to the data by applying the one or more trained machine learning models to generate links between the atomic knowledge units based on the one or more attributes;

processing the generated links by a model ensemble implementing an ensemble learning to establish contextual links between the data from the plurality of data sources;

generating a linked data set based on an analysis of the contextual links; and

generating the one or more insights related to the building construction project based on the linked data set and the one or more construction objectives, wherein the construction objectives comprise at least one of schedule optimization of the building construction project, cost optimization of the building construction project and carbon footprint optimization of the building construction project, and the one or more insights identify building construction project actions that contribute to achievement of at least one of the construction objectives.

2. The method of claim 1 , further comprising:

analyzing the linked data set based on the one or more construction objectives,

wherein the generated one or more insights correspond to situational and contextual insights related to the building construction project and are based on the analysis of the linked data set in view of the one or more construction objectives.

3. The method of claim 1 , wherein applying the one or more trained machine learning models further comprising:

applying attribute-based machine learning models on the atomic knowledge units; and

applying non-attribute-based machine learning models trained on construction data on the atomic knowledge units to generate links between the atomic knowledge units.

4. The method of claim 1 , wherein decomposing the received data into the decomposed data set further comprising:

breaking down the received data into smaller units of relevant data and irrelevant data; and

removing the irrelevant data from the decomposed data set.

5. The method of claim 1 , wherein parsing and analyzing the decomposed data set based on the set of attribute analyzers comprises analyzing the decomposed data set based on one or more of a semantic analyzer, a temporal analyzer, a resource analyzer, an intent analyzer, and a location parser.

6. The method of claim 1 , further comprising storing the atomic knowledge units in a multi-dimensional data format including one or more of a multi-dimensional cube representation, a vector representation, word embeddings, a semantic representation, and a linked graph structure data representation.

7. The method of claim 1 , wherein analyzing the atomic knowledge units to generate links between the atomic knowledge units further comprising processing, by a Natural Language Processing (NLP) module, the atomic knowledge units to perform a semantic feature analysis of the atomic knowledge units.

8. The method of claim 7 , further comprising:

analyzing, by a NLP classifier, the processed atomic knowledge units to perform text classification by assigning a set of tags to one or more portions of the atomic knowledge units; and

generating links between atomic knowledge units based on a relatability of the set of tags.

9. The method of claim 1 , wherein processing the generated links by the model ensemble further comprising:

determining a score associated with each generated link; and

correlating the generated links based on the score and one or more construction objectives to establish the contextual links between data from the plurality of data sources.

10. A system for generating one or more insights related to a building construction project having one or more construction objectives by establishing and generating associative contextual links between data from a plurality of data sources in a computing environment, said system comprising:

an Autolink Module having a controller configured to:

obtain, via a network, the data from the plurality of data sources based on a request received from a client computer for the one or more insights associated with the building construction project, wherein the plurality of data sources corresponds to disparate systems generating unlinked data;

generate a decomposed data set comprising a relevant data set from the obtained data;

parse and analyze the decomposed data set based on a set of attribute analyzers to associate one or more attributes to the decomposed data set;

determine an intent of data from the decomposed data set associated with each of the plurality of data sources by applying a natural language parser;

generate a semantic graph of the decomposed data set based on the intent of data to evaluate data relatability between the decomposed data set;

generate atomic knowledge units corresponding to the data based on the parsed decomposed data set and the semantic graph, wherein the atomic knowledge units represent the data in an organized data format;

train one or more machine learning models using training data, wherein the training data includes data relating to the building construction project and one or more user inputs;

analyze the atomic knowledge units corresponding to the received data by applying one or more trained machine learning models to generate links between the atomic knowledge units based on the one or more attributes;

process the generated links by a model ensemble implementing an ensemble learning to establish contextual links between the data from the plurality of data sources;

generate a linked data set based on an analysis of the contextual links; and

generate the one or more insights related to the building construction project based on the linked data set and the one or more construction objectives, wherein the construction objectives comprise at least one of schedule optimization of the building construction project, cost optimization of the building construction project and carbon footprint optimization of the building construction project, and the one or more insights identify building construction project actions that contribute to achievement of at least one of the construction objectives.

11. The system of claim 10 , further comprising an Insights Module configured to:

analyze the linked data set based on the one or more construction objectives,

wherein the generated one or more insights correspond to situational and contextual insights related to the building construction project and are based on the analysis of the linked data set in view of the one or more construction objectives.

12. The system of claim 10 , wherein the controller is further configured to:

apply attribute-based machine learning models on the atomic knowledge units; and

apply non-attribute-based machine learning models trained on construction data on the atomic knowledge units to generate links between the atomic knowledge units.

13. The system of claim 10 , the controller is further configured to:

break down the received data into smaller units of relevant data and irrelevant data; and

remove the irrelevant data from the received data to generate the decomposed data set.

14. The system of claim 10 , further comprising an Atomic Knowledge Units Generator configured to parse and analyze the decomposed data set, the Atomic Knowledge Units Generator comprises a semantic analyzer, a temporal analyzer, a resource analyzer, an intent analyzer, and a location parser.

15. The system of claim 10 , further comprising a Multi-dimensional Data Module configured to store the atomic knowledge units in a multi-dimensional data format including one or more of a multi-dimensional cube representation, a vector representation, word embeddings, a semantic representation, and a linked graph structure data representation.

16. The system of claim 10 , further comprising a Natural Language Processing (NLP) Module configured to perform a semantic feature analysis of the atomic knowledge units.

17. The system of claim 16 , further comprising an NLP Classifier configured to perform text classification on processed atomic knowledge units by assigning a set of tags to one or more portions of the atomic knowledge units, wherein the links between atomic knowledge units are generated based on the set of tags.

18. A non-transitory computer-readable storage medium, having stored thereon a computer-executable program for generating one or more insights related to a building construction project having one or more construction objectives, which, when executed by at least one processor, causes the at least one processor to:

obtain, via a network, data from a plurality of data sources based on a request received from a client computer for the one or more insights associated with the building construction project, wherein the plurality of data sources corresponds to disparate systems generating unlinked data;

generate a decomposed data set comprising a relevant data set from the obtained data;

parse and analyze the decomposed data set based on a set of attribute analyzers to associate one or more attributes to the decomposed data set;

determine an intent of data from the decomposed data set associated with each of the plurality of data sources by applying a natural language parser;

generate a semantic graph of the decomposed data set based on the intent of data to evaluate data relatability between the decomposed data set;

generate atomic knowledge units corresponding to the data based on the parsed decomposed data set and the semantic graph, wherein the atomic knowledge units represent the data in an organized data format;

train one or more machine learning models using training data, wherein the training data includes data relating to the building construction project and one or more user inputs;

analyze the atomic knowledge units corresponding to the received data by applying one or more trained machine learning models to generate links between the atomic knowledge units based on the one or more attributes;

process the generated links by a model ensemble implementing an ensemble learning to establish contextual links between the data from the plurality of data sources;

generate a linked data set based on an analysis of the contextual links; and

generate the one or more insights related to the building construction project based on the linked data set and the one or more construction objectives, wherein the construction objectives comprise at least one of schedule optimization of the building construction project, cost optimization of the building construction project and carbon footprint optimization of the building construction project, and the one or more insights identify building construction project actions that contribute to achievement of at least one of the construction objectives.

Assignments (1)

References Cited (135)

Cited By (1)

US 12,579,369