Ontology-augmented interface

Abstract

Provided is a process including obtaining a set of natural-language text documents that discuss a topic, the set of documents containing different states of knowledge about the topic at different times. The process includes selecting an ontology from among a plurality of ontologies that correspond to different domains of knowledge, the selection being based on the ontology corresponding to a domain of knowledge including the topic. The process includes identifying concepts discussed in the documents using the ontology and detecting changes in at least some of the concepts over time based on differences between discussion of the concepts in documents authored at different times. The process includes updating natural language instructions on the topic based on the detected changes in the concepts and storing the updated natural language instructions in memory.

What is claimed is: 
     
         1 . A method, comprising:
 obtaining, with a computer system, via a network, a first message from a client computing device, wherein the first message identifies a natural-language text document;   sending, with the computer system, a user interface (UI) to the client computing device via a response to the first message, the UI comprising:
 a display of the natural-language text document that comprises n-grams associated with a set of ontologies; 
 a visual indicator indicating an ontology-linked n-gram of the natural-language text document is mapped to an entry of an ontology; 
 a UI element that, when interacted with, causes the client computing device to send a second message to the computer system, the second message comprising the ontology-linked n-gram; 
   obtaining, with the computer system, the second message from the client computing device after an interaction with the UI to update the ontology;   updating, with the computer system, the entry mapped to the ontology-linked n-gram; and   updating, with the computer system, a decision tree based on the update to the ontology.   
     
     
         2 . The method of  claim 1 , wherein the natural-language text document is a first version of the natural-language text document, the method further comprising:
 obtaining a previous version of the natural-language text document;   determining a text difference between the first version of the natural-language text document and the previous version of the natural-language text document, wherein the text difference comprises the ontology-linked n-gram;   determining a domain associated with the text difference based on the ontology; and   sending a third message to the client computing device to visually indicate the domain in the UI.   
     
     
         3 . The method of  claim 2 , wherein the visual indicator is a first visual indicator, further comprising visually indicating the text difference using a second visual indicator, wherein the second visual indicator surrounds the ontology-linked n-gram. 
     
     
         4 . The method of  claim 1 , wherein the visual indicator is a first visual indicator, and wherein the entry is a first vertex of the ontology, the method further comprising:
 obtaining a third message;   updating the ontology based on the third message by adding a second vertex to the ontology, wherein the second vertex maps to a second ontology-linked n-gram; and   updating the UI by adding a second visual indicator indicating the second ontology-linked n-gram of the natural-language text document.   
     
     
         5 . The method of  claim 1 , wherein updating the decision tree comprises:
 determining whether a domain associated with the ontology is listed in a set of decision-impacting domains;   in response to a determination that the domain is listed in the set of decision-impacting domains, determining a set of decision tree nodes associated with the domain; and   send a third message indicating the set of decision tree nodes to the client computing device.   
     
     
         6 . The method of  claim 1 , further comprising:
 obtaining a third message indicating a request to access a set of workflow operations associated with a structured data document, wherein sending the UI comprises:   sending a set of values causing the UI to display a set of shapes representing the set of workflow operations, wherein:
 a first shape of the set of shapes corresponds with a data ingestion operation of the set of workflow operations; 
 a second shape of the set of shapes corresponds with a data ingestion source, wherein the second shape is connected with the first shape on the UI; 
 a third shape of the set of shapes corresponds with a data processing operation of the set of workflow operations, wherein a parameter of the data processing operation comprises an identifier of a neural network model; 
 a fourth shape of the set of shapes is connected with both the first shape and the third shape; and 
 the set of ontologies is displayed in an interactive UI element as an input or output for at least one shape of the set of shapes; 
   obtaining an update message to the set of workflow operations based on a reconfiguration of the set of shapes; and   re-compiling program code corresponding with the structured data document to generate an intermediate representation of the program code in response to receiving the update message.   
     
     
         7 . The method of  claim 6 , further comprising generating an updated version of the structured data document after re-compiling the program code, wherein the updated structured data document comprises:
 a first set of values enclosed in a first pair of brackets, the first set of values comprising an identifier of a data ingestion source; and   a second set of values enclosed in a second pair of brackets, the second set of values comprising an identifier of the neural network model.   
     
     
         8 . A computer-implemented method, comprising:
 obtaining, with a computer system, a set of natural-language text documents that discuss a topic, the set of documents containing different states of knowledge about the topic at different times;   selecting, with the computer system, an ontology from among a plurality of ontologies that correspond to different domains of knowledge, the selection being based on the ontology corresponding to a domain of knowledge including the topic;   identifying, with the computer system, using the ontology, concepts discussed in the documents;   detecting, with the computer system, changes in at least some of the concepts over time based on differences between discussion of the concepts in documents authored at different times;   updating, with the computer system, natural language instructions on the topic based on the detected changes in the concepts; and   storing, with the computer system, the updated natural language instructions in memory.   
     
     
         9 . The method of  claim 8 , wherein updating the natural language instructions comprises:
 causing a user interface to be displayed, the user interface indicating both the updated natural language instructions and changes in the natural language instructions relative to a prior version of the natural language instructions.   
     
     
         10 . The method of  claim 9 , wherein:
 the user interface includes links to, or displays at least some text from, documents among the set of natural-language text documents indicative of corresponding changes to the instructions.   
     
     
         11 . The method of  claim 9 , wherein:
 the natural language instructions comprise a flow chart, a decision tree, a manual, a regulation, or a policy.