System and method of keyword sensitive semantic search scoring for hierarchically related artificial intelligence productivity tool-enablable application capabilities for a user query input at an information handling system
An information handling system for an on-the-box artificial intelligence productivity tool may comprise a memory storing descriptions of capabilities associated with software applications in a decision tree with each capability and a generated capability intent value stored as a capability node grouped under one of several branches in parent-child relationships, and a hardware processor executing machine readable code instructions to generate a query input intent value from a user query input requesting action by one of the software applications, comparing the capability intent values of the capability nodes along a branch in the decision tree with a weighted by a TF-IDF comparison between the user query input and each of the descriptions of capabilities to identify a best match capability node having a highest TF-IDF weighted cosine semantic similarity search score, and executing an associated best match capability for a first software application having the best match capability node.
1 . An information handling system executing computer readable code instructions for an on the box (OTB) artificial intelligence (AI) productivity tool comprising:
a natural language capabilities database memory to store natural language descriptions of capabilities associated with each of a plurality of AI productivity tool-enablable software applications executing on the information handling system in a capabilities decision tree with each capability stored under a plurality of branches as a capability node grouped under a selected branch of the capabilities decision tree according to logical topics in hierarchical parent-child relationships, wherein metadata for each capability node identifies a child capability node or parent capability node of the capability node;
a hardware processor executing computer-readable program code instructions of the OTB AI productivity tool to generate capability intent values from the natural language descriptions of the capabilities and storing the capability intent values with the capability nodes;
the hardware processor executing computer-readable program code instructions to generate a query input intent value from a user query input received via text or audio requesting an action by one of the plurality of AI productivity tool-enablable software applications;
the hardware processor executing computer-readable program code instructions to perform a text frequency-inverted document frequency (TF-IDF) weighted cosine semantic similarity search semantically comparing the capability intent values of the capability nodes along the selected branch from the plurality of branches in the capabilities decision tree with the query input intent value, as weighted by a TF-IDF comparison between natural language of the user query input and each of the natural language descriptions of capabilities of the capability nodes to identify a best match capability node having a highest TF-IDF weighted cosine semantic similarity search score with the query input intent value; and
the hardware processor executing computer-readable program code instructions for a first AI productivity tool-enablable software application having the best match capability node to execute an associated best match capability in response to the user query input.
2 . The information handling system of claim 1 , wherein the TF-IDF comparison is conducted with computer readable code instructions for a best match 25 (BM25) algorithm.
3 . The information handling system of claim 1 , wherein the TF-IDF weighted cosine semantic similarity search includes a parent score and TF-IDF weighted cosine semantic similarity search that generates, for each child capability node in the branch, a parent and TF-IDF weighted cosine similarity search score that is weighted by the TF-IDF weighted cosine similarity search score determined for the parent capability node.
4 . The information handling system of claim 1 , wherein the hardware processor executing computer-readable program code instructions to perform the TF-IDF weighted cosine semantic similarity search compares the capability intent values of plural capability nodes along a first level of the capabilities decision tree to determine a parent capability node having a highest TF-IDF weighted cosine semantic similarity search score on the first level of the capabilities tree for selecting the selected branch from the plurality of branches in the capabilities decision tree having child capability nodes to be further analyzed by the TF-IDF weighted cosine semantic similarity search to identify the best match capability node among the capability nodes along the selected branch.
5 . The information handling system of claim 1 further comprising:
the hardware processor executing computer-readable program code instructions of the first AI productivity tool-enablable software application to perform the best match capability to automatically execute changes to or update one or more local software applications in response to the user query input.
6 . The information handling system of claim 1 , wherein the capability intent values are generated by execution of code instructions for a text embedding algorithm and mathematically represent semantic meaning for words or phrases within the natural language descriptions for the capabilities for correlation with the query intent input value generated from the user query input.
7 . The information handling system of claim 1 , wherein the TF-IDF weighted cosine semantic similarity search determines a degree of angular similarity between vector values for the capability intent values and the query input intent value that mathematically represent one or more phrases within the natural language descriptions for the capabilities and natural language of the user query input as weighted by the TF-IDF comparison.
8 . The information handling system of claim 1 , wherein the TF-IDF comparison is conducted with computer readable code instructions for a best match 25 (BM25) algorithm.
9 . A method for executing computer readable code instructions of an on the box (OTB) artificial intelligence (AI) productivity tool at an information handling system to respond to a user query input comprising:
storing, in a natural language capabilities database memory, the natural language descriptions of capabilities for a plurality of AI productivity tool-enablable software applications executing on the information handling system, and capability intent values generated from the natural language descriptions in capability nodes in a capabilities decision tree, where each capability node is grouped under a branch of a plurality of branches in the capabilities decision tree according to logical topics in hierarchical parent-child relationships and where metadata for each of those capability nodes identifies a child capability node or parent capability node of each of those capability nodes;
generating, via a hardware processor executing machine readable code instructions of the OTB AI productivity tool, a query input intent value from the user query input received via text or audio requesting an action by one of the plurality of AI productivity tool-enablable software applications;
performing, via a hardware processor, a text frequency-inverted document frequency (TF-IDF) weighted semantic similarity search comparing the capability intent values of the each of the capability nodes along the branch of the plurality of branches in the capabilities decision tree with the query input intent value and weighted by a TF-IDF comparison between the user query input and the natural language descriptions of the capability nodes along the branch of the plurality of branches in the capabilities decision tree to identify a best match capability node among the capability nodes along the branch having a highest TF-IDF weighted semantic similarity search score with the user query input; and
executing computer-readable program code instructions for a first AI productivity tool-enablable software application having the best match capability node to execute an associated best match capability in response to the user query input.
10 . The method of claim 9 wherein the TF-IDF weighted semantic similarity search includes a parent score and TF-IDF weighted semantic similarity search that generates, for each child capability node in the branch, a parent and TF-IDF weighted similarity search score that is weighted by the TF-IDF weighted similarity search score determined for the parent capability node.
11 . The method of claim 9 , wherein the capability intent values are generated by execution of code instructions for a text embedding algorithm and mathematically represent semantic meaning for words or phrases within the natural language descriptions for the capabilities for correlation with the query input intent value generated from natural language of the user query input.
12 . The method of claim 9 , wherein computer-readable program code instructions of a latent semantic analysis text embedding algorithm are executed via the hardware processor for generating the capability intent values and the query input intent value.
13 . The method of claim 9 , wherein the hardware processor executing computer-readable program code instructions to perform the TF-IDF weighted semantic similarity search compares the capability intent values of plural capability nodes along a first level of the capabilities decision tree to determine a parent capability node having a highest TF-IDF weighted semantic similarity search score on the first level of the capabilities tree for selecting the branch of the plurality of branches in the capabilities decision tree having child capability nodes to be further analyzed by the TF-IDF weighted semantic similarity search to identify the best match capability node among the capability nodes along the branch.
14 . The method of claim 9 , wherein a cosine semantic search machine learning algorithm generates the TF-IDF weighted semantic similarity search score for each of the capabilities by performing a cosine semantic similarity search comparing a degree of angular similarity between vector values for the capability intent values in the capability nodes of the branch of the plurality of branches in the capabilities decision tree with the query input intent value and weighting the cosine semantic similarity search with the TF-IDF comparison of terms in the user query input with the natural language descriptions of each of the capability nodes.
15 . The method of claim 9 further comprising:
executing computer-readable program code instructions of the first AI productivity tool-enablable software application, via the hardware processor, to perform the best match capability to optimize settings for a hardware component of the information handling system in response to the user query input.
16 . An information handling system executing computer readable code instructions for an on the box (OTB) artificial intelligence (AI) productivity tool comprising:
a natural language capabilities database memory to store natural language descriptions of capabilities associated with each of a plurality of AI productivity tool-enablable software applications executing on the information handling system, capability intent values generated from the natural language descriptions for each capability, and a capability identification in capability nodes in a capabilities decision tree with each of the capability nodes grouped under a branch of a plurality of branches in the capabilities decision tree according to logical topics in hierarchical parent-child relationships, wherein metadata for each of the capability nodes identifies a child capability node or parent capability node of each of those capability nodes;
the hardware processor executing computer-readable program code instructions to generate a query input intent value from a user query input received via text or audio requesting an action by one of the plurality of AI productivity tool-enablable software applications;
the hardware processor executing computer-readable code instructions for performing a text frequency-inverted document frequency (TF-IDF) comparison between natural language of the user query input and each of the natural language descriptions of the capabilities along the branch of the capabilities decision tree for TF-IDF similarity values; and
the hardware processor executing computer-readable program code instructions to perform a TF-IDF weighted cosine semantic similarity search comparing the capability intent values of the capability nodes along the branch of the plurality of branches in the capabilities decision tree and weighted by the TF-IDF similarity values to identify a best match capability node having a highest TF-IDF weighted cosine semantic similarity search score with the query input intent value; and
the hardware processor executing computer-readable program code instructions for a first AI productivity tool-enablable software application having the best match capability node to execute an associated best match capability in response to the user query input.
17 . The information handling system of claim 16 , wherein the capability intent values are generated by execution of code instructions for a text embedding algorithm and mathematically represent semantic meaning for words or phrases within the natural language descriptions for the gathered capabilities for correlation with the natural language of the query input intent value generated from the natural language of the user query input.
18 . The information handling system of claim 16 , wherein the hardware processor executes computer-readable program code instructions of a cosine semantic search machine learning algorithm for generating the TF-IDF weighted semantic similarity search score for each of the capabilities by comparing the capability intent values in the branch of the plurality of branches in the capabilities decision tree to the query input intent value.
19 . The information handling system of claim 16 , wherein the hardware processor executing computer-readable program code instructions to perform the TF-IDF weighted semantic similarity search compares the capability intent values of plural capability nodes along a first level of the capabilities decision tree to determine a parent capability node having a highest TF-IDF weighted semantic similarity search score on the first level of the capabilities tree for selecting the branch of the plurality of branches in the capabilities decision tree having child capability nodes to be further analyzed by the TF-IDF weighted semantic similarity search to identify the best match capability node among the capability nodes along the branch.
20 . The information handling system of claim 16 further comprising:
the hardware processor executing computer-readable program code instructions of the first AI productivity tool-enablable software application to perform the best match capability to provide hardware component adjustment for a hardware component of the information handling system in response to the user query input.