Feature recommendation based on user-generated content

1. A computing system comprising:

a memory storing a topic modeling algorithm; and

a processor configured to

receive, via a user interface, a search query associated with an object,

retrieve data structures comprising user-generated content that describes the object based on the received search query,

execute a machine learning model on the user-generated content within the data structures to identify a first subset of data structures with user-generated content corresponding to feature-related data of the object therein and a second subset of data structures with user-generated content corresponding to non-feature related data therein based on patterns of words in the data structures,

divide the first subset of data structures with the feature-related data of the object identified by the machine learning model into a plurality of clusters corresponding to a plurality of different features of the object, respectively, via execution of a topic modeling algorithm,

rank the plurality of different features based on how many records are included in each cluster to identify a highest ranking feature,

determine a recommended upgrade to the highest ranking feature of the object from among the plurality of different features based on the execution of the topic modeling algorithm on data structures within a cluster from among the plurality of clusters which corresponds to the highest ranking feature; and

output an identifier of the recommended upgrade to the feature of the object via the user interface.

2. The computing system of claim 1 , wherein the processor is further configured to convert the user-generated content into the data structures, where each data structure comprises a predefined format which includes values extracted from the user-generated content stored in predefined fields corresponding to object attributes.

3. The computing system of claim 1 , wherein the processor is configured to determine the highest priority feature to be added to the object from among the plurality of different features based on weights added to the plurality of different features by the topic modeling algorithm.

4. The computing system of claim 1 , wherein the search query comprises one or more of a name of the object, a name of a model of the object, and a description of the object, input via one or more fields of the user interface.

5. The computing system of claim 1 , wherein the processor is further configured to modify the search query to include an additional descriptive term associated with the object.

6. The computing system of claim 1 , wherein the processor is configured to identify a new feature to add that is not present in the object based on execution of the topic modeling algorithm on the data structures within the cluster.

7. The computing system of claim 1 , wherein the processor is configured to generate a summary description from the data structures within the cluster from among the plurality of clusters, and output the summary description with the recommended upgrade to the highest ranking feature via the user interface.

8. The computing system of claim 1 , wherein the processor is further configured to export a description of the recommended upgrade to the highest ranking feature into an external software application used for designing the object.

9. A method comprising:

receiving, via a user interface, a search query associated with an object;

retrieving data structures comprising user-generated content that describes the object based on the received search query;

executing a machine learning model on the user-generated content within the data structures to identify a first subset of data structures with user-generated content corresponding to feature-related data of the object therein and a second subset of data structures with user-generated content corresponding to non-feature related data therein based on patterns of words in the data structures;

dividing the first subset of data structures with the feature-related data of the object identified by the machine learning model into a plurality of clusters corresponding to a plurality of different features of the object, respectively, via execution of a topic modeling algorithm;

ranking the plurality of different features based on how many records are included in each cluster to identify a highest ranking feature;

determining a recommended upgrade to the highest ranking feature of the object from among the plurality of different features based on the execution of a topic modeling algorithm on data structures within a cluster from among the plurality of clusters which corresponds to the highest ranking feature; and

outputting an identifier of the recommended upgrade to the feature of the object via the user interface.

10. The method of claim 9 , further comprising converting the user-generated content into the data structures, where each data structure comprises a predefined format which includes values extracted from the user-generated content stored in predefined fields corresponding to object attributes.

11. The method of claim 9 , wherein the determining comprises determining the highest priority feature to be added to the object from among the plurality of different features based on weights added to the plurality of different features by the topic modeling algorithm.

12. The method of claim 9 , wherein the search query comprises one or more of a name of the object, a name of a model of the object, and a description of the object, input via one or more fields of the user interface.

13. The method of claim 9 , further comprising modifying the search query to include an additional descriptive term associated with the object.

14. The method of claim 9 , wherein the executing comprises identifying a new feature to add that is not present in the object based on execution of the topic modeling algorithm on the data structures within the cluster.

15. The method of claim 9 , wherein the determining further comprises generating a summary description from the data structures within the cluster from among the plurality of clusters, and outputting the summary description with the recommended upgrade to the highest ranking feature via the user interface.

16. A non-transitory computer-readable medium comprising instructions which when executed by a processor cause a computer to perform a method comprising:

receiving, via a user interface, a search query associated with an object;

retrieving data structures comprising user-generated content that describes the object based on the received search query;

executing a machine learning model on the user-generated content within the data structures to identify a first subset of data structures with user-generated content corresponding to feature-related data of the object therein and a second subset of data structures with user-generated content corresponding to non-feature related data therein based on patterns of words in the data structures;

dividing the first subset of data structures with the feature-related data of the object identified by the machine learning model into a plurality of clusters corresponding to a plurality of different features of the object, respectively, via execution of a topic modeling algorithm;

ranking the plurality of different features based on how many records are included in each cluster to identify a highest ranking feature;

determining a recommended upgrade to the highest ranking feature of the object from among the plurality of different features based on the execution of a topic modeling algorithm on data structures within a cluster from among the plurality of clusters which corresponds to the highest ranking feature; and

outputting an identifier of the recommended upgrade to the feature of the object via the user interface.

17. The non-transitory computer-readable of claim 16 , wherein the method further comprises converting the user-generated content retrieved into the data structures, where each data structure comprises a predefined format which includes values extracted from the user-generated content stored in predefined fields corresponding to object attributes.

18. The non-transitory computer-readable of claim 16 , wherein the determining comprises determining the highest priority feature from among the plurality of different features based on weights added to the plurality of different features by the topic modeling algorithm.

19. The non-transitory computer-readable of claim 16 , wherein the executing comprises identifying a new feature to add that is not present in the object based on execution of the topic modeling algorithm on the data structures within the cluster.