Methods and apparatus to facilitate generation of database queries

Abstract

Methods and apparatus to facilitate generation of database queries are disclosed. An example apparatus includes a generator to generate a global importance tensor. The global importance tensor based on a knowledge graph representative of information stored in a database. The knowledge graph includes objects and connections between the objects. The global importance tensor includes importance values for different types of the connections between the objects. The example apparatus further includes an importance adaptation analyzer to generate a session importance tensor based on the global importance tensor and a user query, and a user interface to provide a suggested query to a user based on the session importance tensor.

1 . (canceled) 
     
     
         2 - 25 . (canceled) 
     
     
         26 . At least one non-transitory computer readable medium comprising instructions to cause processor circuitry to at least:
 access first data associated with a database;   generate a knowledge graph representative of relationships between data items included in the first data, the data items represented by vertices in the knowledge graph, the relationships between the data items represented by edges in the knowledge graph, different ones of the edges having different types;   analyze the knowledge graph independent of a first user query to generate second data, the second data corresponding to statistics associated with at least one of the vertices, the edges, or the types of edges in the knowledge graph, the second data distinct from the first data and distinct from the knowledge graph;   identify information represented in the knowledge graph based on the first user query from the user obtained via a user interface; and   prompt the user for additional input based on analysis of the first user query and the second data, the additional input to drive a second user query.   
     
     
         27 . The at least one computer readable medium of  claim 26 , wherein the instructions cause the processor circuitry to update the second data based on at least one of (i) a schedule or (ii) a change in the data items in the database represented in the knowledge graph. 
     
     
         28 . The at least one computer readable medium of  claim 26 , wherein the second data is a first importance tensor containing importance values for the different types of the edges, the instructions to cause the processor circuitry to generate a second importance tensor based on the first importance tensor and based on the first user query, the additional input based on the second importance tensor. 
     
     
         29 . The at least one computer readable medium of  claim 28 , wherein the instructions cause the processor circuitry to generate the second importance tensor by:
 defining the second importance tensor as corresponding to the first importance tensor;   calculating a loss function for the second importance tensor based on ones of the data items and associated ones of the edges identified in the first user query; and   updating the second importance tensor by a negative gradient of the loss function.   
     
     
         30 . The at least one computer readable medium of  claim 28 , wherein the instructions cause the processor circuitry to generate the importance values based on at least one of a number of respective ones of different types of the data items in the knowledge graph, a number of response ones of the different type of the edges in the knowledge graph, an average number of the edges associated with respective ones of the different types of the data items, or a standard deviation of a number of the edges associated with ones of the data items corresponding to particular ones of the different types of the data items. 
     
     
         31 . The at least one computer readable medium of  claim 28 , wherein the instructions cause the processor circuitry to:
 calculate a first weight for a first one of the data items;   calculate a second weight for a second one of the data items;   calculate a third weight for a first one of the edges between the first and second data items; and   calculate a first one of the importance values in the first importance tensor by multiplying the first, second, and third weights.   
     
     
         32 . The at least one computer readable medium of  claim 31 , wherein the first data item corresponds to a first type of data item, the instructions to cause the processor circuitry to calculate the first weight by:
 calculating a first number of the data items in the knowledge graph corresponding to the first type;   calculating an average number of the edges associated with each of the data items in the knowledge graph corresponding to the first type; and   calculating a standard deviation of a number of the edges associated with each of the data items corresponding to the first type.   
     
     
         33 . The at least one computer readable medium of  claim 31 , wherein the first edge corresponds to a first edge type, the instructions to cause the processor circuitry to calculate the third weight by:
 calculating a number of the edge in the knowledge graph corresponding to the first edge type; and   normalizing the number relative to a total number of the edge in the knowledge graph.   
     
     
         34 . The at least one computer readable medium of  claim 26 , wherein the instructions cause the processor circuitry to prompt the user for additional input by providing the second user query for selection. 
     
     
         35 . A system comprising:
 memory;   machine readable instructions; and   processor circuitry to execute the instructions to:
 generate a knowledge graph representing of relationships between data items included in first data, the first data associated with a database, the data items represented by vertices in the knowledge graph, the relationships between the data items represented by edges in the knowledge graph, different edges having different types; 
 analyze the knowledge graph prior to receipt of a first user query to generate second data, the second data corresponding to statistics associated with at least one of the vertices, the edges, or the types of edges in the knowledge graph, the second data distinct from the first data and distinct from the knowledge graph; 
 identify information represented in the knowledge graph based on the first user query from the user obtained via a user interface; and 
 prompt the user for additional input based on analysis of the first user query and the second data, the additional input to drive a second user query. 
   
     
     
         36 . The system of  claim 35 , wherein the processor circuitry is to update the second data based on at least one of (i) a schedule or (ii) a change in the data items in the database represented in the knowledge graph. 
     
     
         37 . The system of  claim 35 , wherein the second data is a first importance tensor containing importance values for the different types of the edges, the processor circuitry to generate a second importance tensor based on the first importance tensor and based on the first user query, the additional input based on the second importance tensor. 
     
     
         38 . The system of  claim 37 , wherein the processor circuitry is to generate the second importance tensor by:
 defining the second importance tensor as corresponding to the first importance tensor;   calculating a loss function for the second importance tensor based on ones of the data items and associated ones of the edges identified in the first user query; and   updating the second importance tensor by a negative gradient of the loss function.   
     
     
         39 . The system of  claim 37 , wherein the processor circuitry is to calculate the importance values based on at least one of a number of respective ones of different types of the data items in the knowledge graph, a number of response ones of the different type of the edges in the knowledge graph, an average number of the edges associated with respective ones of the different types of the data items, or a standard deviation of a number of the edges associated with ones of the data items corresponding to particular ones of the different types of the data items. 
     
     
         40 . The system of  claim 37 , wherein the processor circuitry is to:
 calculate a first weight for a first one of the data items;   calculate a second weight for a second one of the data items;   calculate a third weight for a first one of the edges between the first and second data items; and   calculate a first one of the importance values in the first importance tensor by multiplying the first, second, and third weights.   
     
     
         41 . The system of  claim 40 , wherein the first data item corresponds to a first type of data item, the processor circuitry to calculate the first weight by:
 calculating a first number of the data items in the knowledge graph corresponding to the first type;   calculating an average number of the edges associated with each of the data items in the knowledge graph corresponding to the first type; and   calculating a standard deviation of a number of the edges associated with each of the data items corresponding to the first type.   
     
     
         42 . The system of  claim 40 , wherein the first edge corresponds to a first edge type, the processor circuitry to calculate the third weight by:
 calculating a number of the edge in the knowledge graph corresponding to the first edge type; and   normalizing the number relative to a total number of the edge in the knowledge graph.   
     
     
         43 . The system of  claim 35 , wherein the processor circuitry is to prompt the user for additional input by providing the second user query for selection. 
     
     
         44 . A method comprising:
 accessing first data associated with a database;   generating a knowledge graph representing relationships between data items included in the first data, the data items represented by vertices in the knowledge graph, the relationships between the data items represented by edges in the knowledge graph, different edges having different types;   analyzing, by performing an operation with processor circuitry, the knowledge graph independent of a first user query to generate second data, the second data corresponding to statistics associated with at least one of the vertices, the edges, or the types of edges in the knowledge graph, the second data distinct from the first data and distinct from the knowledge graph;   identifying information represented in the knowledge graph based on the first user query from the user obtained via a user interface; and   prompting the user for additional input based on analysis of the first user query and the second data, the additional input to drive a second user query.   
     
     
         45 . The method of  claim 44 , further including updating the second data based on at least one of (i) a schedule or (ii) a change in the data items in the database represented in the knowledge graph. 
     
     
         46 . The method of  claim 44 , wherein the second data is a first importance tensor containing importance values for the different types of the edges, the method further including generating a second importance tensor based on the first importance tensor and based on the first user query, the additional input based on the second importance tensor. 
     
     
         47 . The method of  claim 46 , wherein the generating of the second importance tensor includes:
 defining the second importance tensor as corresponding to the first importance tensor;   calculating a loss function for the second importance tensor based on ones of the data items and associated ones of the edges identified in the first user query; and   updating the second importance tensor by a negative gradient of the loss function.   
     
     
         48 . The method of  claim 46 , further including calculating the importance values based on at least one of a number of respective ones of different types of the data items in the knowledge graph, a number of response ones of the different type of the edges in the knowledge graph, an average number of the edges associated with respective ones of the different types of the data items, or a standard deviation of a number of the edges associated with ones of the data items corresponding to particular ones of the different types of the data items. 
     
     
         49 . The method of  claim 46 , further including to:
 calculating a first weight for a first one of the data items;   calculating a second weight for a second one of the data items;   calculating a third weight for a first one of the edges between the first and second data items; and   calculating a first one of the importance values in the first importance tensor by multiplying the first, second, and third weights.   
     
     
         50 . The method of  claim 49 , wherein the first data item corresponds to a first type of data item, the calculating of the first weight including:
 calculating a first number of the data items in the knowledge graph corresponding to the first type;   calculating an average number of the edges associated with each of the data items in the knowledge graph corresponding to the first type; and   calculating a standard deviation of a number of the edges associated with each of the data items corresponding to the first type.   
     
     
         51 . The method of  claim 49 , wherein the first edge corresponds to a first edge type, the calculating of the third weight including:
 calculating a number of the edge in the knowledge graph corresponding to the first edge type; and   normalizing the number relative to a total number of the edge in the knowledge graph.   
     
     
         52 . The method of  claim 44 , wherein the prompting of the user for additional input includes providing the second user query for selection. 
     
     
         53 . The method of  claim 44 , wherein the identifying of the information includes identifying information of interest to the user.