Systems, methods, and apparatuses for implementing predictive query interface as a cloud service
First Claim
1. A method in a host organization, the method comprising:
- exposing an interface to client devices operating remotely from the host organization, wherein the interface is accessible by the client devices via a public Internet;
executing a predictive database at the host organization as an on-demand cloud based service for one or more subscribers;
authenticating one of the client devices by verifying the client device is associated with one of the subscribers and based further on authentication credentials for the respective subscriber;
processing a dataset of columns and rows to generate indices on behalf of the authenticated subscriber, the indices representing probabilistic relationships between the rows and the columns of the dataset, wherein the processing comprises;
(i) processing the dataset by iteratively learning joint probability distributions over the dataset to generate the indices, (ii) periodically determining a predictive quality measure of the indices generated by the processing of the dataset, and (iii) terminating processing of the dataset when the predictive quality measure attains a specified threshold;
receiving a prediction request from the authenticated subscriber via the interface;
executing a query against the indices of the predictive database generated from the dataset;
returning a prediction result of the query to the authenticated subscriber responsive to the prediction request; and
returning a notification with the prediction result indicating processing of the stored dataset has not yet completed or returning a notification with the prediction result indicating the predictive quality measure is below the specified threshold, or returning both with the prediction result.
1 Assignment
0 Petitions
Accused Products
Abstract
Disclosed herein are systems and methods for implementing predictive query interface as a cloud service including means for exposing an interface to client devices operating remotely from a host organization, wherein the interface is accessible by the client devices via a public Internet; executing a predictive database at the host organization as an on-demand cloud based service for one or more subscribers; authenticating one of the client devices by verifying the client device is associated with one of the subscribers and based further on authentication credentials for the respective subscriber; receiving a request from the authenticated subscriber via the interface; executing a predictive query or a latent structure query against indices of the predictive database generated from a dataset of columns and rows on behalf of the authenticated subscriber, the indices representing probabilistic relationships between the rows and the columns of the dataset; and returning a predictive record set to the authenticated subscriber responsive to the request. Other related embodiments are further disclosed.
177 Citations
28 Claims
-
1. A method in a host organization, the method comprising:
-
exposing an interface to client devices operating remotely from the host organization, wherein the interface is accessible by the client devices via a public Internet; executing a predictive database at the host organization as an on-demand cloud based service for one or more subscribers; authenticating one of the client devices by verifying the client device is associated with one of the subscribers and based further on authentication credentials for the respective subscriber; processing a dataset of columns and rows to generate indices on behalf of the authenticated subscriber, the indices representing probabilistic relationships between the rows and the columns of the dataset, wherein the processing comprises;
(i) processing the dataset by iteratively learning joint probability distributions over the dataset to generate the indices, (ii) periodically determining a predictive quality measure of the indices generated by the processing of the dataset, and (iii) terminating processing of the dataset when the predictive quality measure attains a specified threshold;receiving a prediction request from the authenticated subscriber via the interface; executing a query against the indices of the predictive database generated from the dataset; returning a prediction result of the query to the authenticated subscriber responsive to the prediction request; and returning a notification with the prediction result indicating processing of the stored dataset has not yet completed or returning a notification with the prediction result indicating the predictive quality measure is below the specified threshold, or returning both with the prediction result. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21)
-
-
22. A non-transitory computer readable storage medium having instructions stored thereon that, when executed by a processor in a host organization, the instructions cause the host organization to perform operations comprising:
-
exposing an interface to client devices operating remotely from the host organization, wherein the interface is accessible by the client devices via a public Internet; executing a predictive database at the host organization as an on-demand cloud based service for one or more subscribers; authenticating one of the client devices by verifying the client device is associated with one of the subscribers and based further on authentication credentials for the respective subscriber; processing a dataset of columns and rows to generate indices on behalf of the authenticated subscriber, the indices representing probabilistic relationships between the rows and the columns of the dataset, wherein the processing comprises;
(i) processing the dataset by iteratively learning joint probability distributions over the dataset to generate the indices, (ii) periodically determining a predictive quality measure of the indices generated by the processing of the dataset, and (iii) terminating processing of the dataset when the predictive quality measure attains a specified threshold;receiving a prediction request from the authenticated subscriber via the interface; executing a query against the indices of the predictive database generated from the dataset; returning a prediction result of the query to the authenticated subscriber responsive to the prediction request; and returning a notification with the prediction result indicating processing of the stored dataset has not yet completed or returning a notification with the prediction result indicating the predictive quality measure is below the specified threshold, or returning both with the prediction result. - View Dependent Claims (23, 24)
-
-
25. A system to operate within a host organization, the system comprising:
-
a processor to execute instructions stored in memory of the system; a request interface exposed to client devices that operate remotely from the host organization, wherein the request interface is accessible by the client devices via a public Internet; a predictive database to execute as an on-demand cloud based service for one or more subscribers; an authenticator to verify client devices are associated with a subscriber and to further verify authentication credentials for the respective subscriber; an analysis engine to process the dataset and generate the indices representing probabilistic relationships between the rows and the columns of the dataset, wherein the analysis engine to process the dataset comprises the analysis engine being adapted to;
(i) process the dataset by iteratively learning joint probability distributions over the dataset to generate the indices, (ii) periodically determine a predictive quality measure of the indices generated by the processing of the dataset, and (iii) terminate processing of the dataset when the predictive quality measure attains a specified threshold;the request interface to receive as input, a prediction request from the subscriber; one or more application servers to execute a query against indices of the predictive database generated from a dataset of columns and rows on behalf of the subscriber, the indices representing probabilistic relationships between the rows and the columns of the dataset; the request interface to return a prediction result of the query to the subscriber responsive to the prediction request; and the request interface to further return a notification with the prediction result that indicates processing of the stored dataset has not yet completed or to return a notification with the prediction result that indicates the predictive quality measure is below the specified threshold, or to return both with the prediction result. - View Dependent Claims (26, 27, 28)
-
Specification