DIGITAL SURVEY CREATION BY PROVIDING OPTIMIZED SUGGESTED CONTENT
The present disclosure relates to a question recommendation system that intelligently optimizes a survey being created by a user by providing customized suggestions. For example, in one or more embodiments, the question recommendation system provides a suggested question based on questions previous added by a user while creating a survey. In particular, the question recommendation system provides various recommendations to the user to further optimize a survey being created. For instance, the question recommendation system provides recommendations with respect to improving question ordering, question phrasing, and question type as well as recommends removing potentially inefficient questions.
1 . A computer-implemented method, comprising:
receiving, from an administrator client device, a user-generated survey prompt for creating a digital survey;
encoding the user-generated survey prompt into a vector space to generate a semantic label;
accessing a database comprising a plurality of survey graphs, wherein each survey graph represents a previously created and administered digital survey and comprises nodes corresponding to survey prompts;
identifying, using connections of the plurality of survey graphs, a set of nodes with semantic labels corresponding to the semantic label of the user-generated survey prompt;
generating, based on response performance metrics associated with the set of nodes, a rephrased survey prompt comprising a rewording of the user-generated survey prompt; and
providing, for display on the administrator client device, the rephrased survey prompt.
2 . The computer-implemented method of claim 1 , wherein providing the rephrased survey prompt further comprises, providing a visual indicator of a historical performance comparison between the user-generated survey prompt and the rephrased survey prompt.
3 . The computer-implemented method of claim 1 , wherein generating the rephrased survey prompt further comprises:
combining the response performance metrics associated with the set of nodes to calculate aggregated response metrics;
selecting, based on a comparison of the aggregated response metrics, a node from the set of nodes with a favorable response metric; and
generating the rewording of the user-generated survey prompt from wording corresponding to the node from set of nodes with the favorable response metric.
4 . The computer-implemented method of claim 1 , further comprising:
displaying, on the administrator client device, one or more alternative rephrasings indicating an associated historical response effectiveness for the one or more alternative rephrasings;
receiving, from the administrator client device, a selection of an alternative rephrasing from the one or more alternative rephrasings for inclusion in the digital survey; and
generating the rephrased survey prompt based on the alternative rephrasing.
5 . The computer-implemented method of claim 1 , further comprising comparing response performance metrics associated with the set of nodes, wherein the response performance metrics comprise a drop-off rate, a skip rate, or a dwell time.
6 . The computer-implemented method of claim 1 , wherein generating the rephrased survey prompt further comprises clustering the set of nodes by vector distances within a semantic space to identify the semantic labels.
7 . The computer-implemented method of claim 6 , wherein the set of nodes is identified by traversing connections within a first survey graph, and further comprising:
identifying a second survey graph linked to the first survey graph via a node with a matching semantic label; and
identifying additional nodes from the second survey graph to include in the set of nodes.
8 . The computer-implemented method of claim 1 , wherein providing the rephrased survey prompt further comprises providing, to the administrator client device, a user interface element enabling a graphical preview of the rephrased survey prompt alongside the user-generated survey prompt.
9 . The computer-implemented method of claim 1 , further comprising:
providing, to the administrator client device, an option to use the rephrased survey prompt; and
based on receiving an indication of a user interaction with the option to use the rephrased survey prompt, generating a digital survey comprising the rephrased survey prompt.
10 . The computer-implemented method of claim 1 , further comprising:
based on a node of the set of nodes satisfying a survey prompt response metric, selecting survey prompt wording corresponding to the node; and
generating the rephrased survey prompt using the survey prompt wording corresponding to the node.
11 . The computer-implemented method of claim 1 , further comprising:
based on receiving, from the administrator client device, an indication to use the rephrased survey prompt, determining to traverse a survey graph of the plurality of survey graphs to identify an additional suggested survey prompt corresponding to a node of the set of nodes within the survey graph; and
providing, to the administrator client device, an additional option to use the additional suggested survey prompt.
12 . A non-transitory computer-readable medium storing instructions that, when executed by at least one processor, cause a computer system to:
receive, from an administrator client device, a user-generated survey prompt for creating a digital survey;
encode the user-generated survey prompt into a vector space to generate a semantic label;
access a database comprising a plurality of survey graphs, wherein each survey graph represents a previously created and administered digital survey and comprises nodes corresponding to survey prompts;
identify, using connections of the plurality of survey graphs, a set of nodes with semantic labels corresponding to the semantic label of the user-generated survey prompt;
generate, based on response performance metrics associated with the set of nodes, a rephrased survey prompt comprising a rewording of the user-generated survey prompt; and
provide, for display on the administrator client device, the rephrased survey prompt.
13 . The non-transitory computer-readable medium of claim 12 , further comprising instructions that, when executed by the at least one processor, cause the computer system to filter the set of nodes based on a comparison of one or more of the response performance metrics to a response metric threshold.
14 . The non-transitory computer-readable medium of claim 12 , further comprising instructions that, when executed by the at least one processor, cause the computer system to determine whether the rephrased survey prompt satisfies a response metric threshold prior to providing the rephrased survey prompt for display.
15 . The non-transitory computer-readable medium of claim 12 , further comprising instructions that, when executed by the at least one processor, cause the computer system to generate the rephrased survey prompt by selecting the rephrased survey prompt based on a survey prompt response metric associated with a node corresponding to the rephrased survey prompt.
16 . The non-transitory computer-readable medium of claim 12 , further comprising instructions that, when executed by the at least one processor, cause the computer system to:
determine that rewording the user-generated survey prompt to the rephrased survey prompt results in a survey pattern corresponding to a historically successful survey pattern; and
provide an option to use the rephrased survey prompt based on determining that rewording the user-generated survey prompt to the rephrased survey prompt will result in the survey pattern corresponding to the historically successful survey pattern.
17 . A system comprising:
at least one processor; and
at least one non-transitory computer-readable storage medium storing instructions that, when executed by the at least one processor, cause the system to:
receive, from an administrator client device, a user-generated survey prompt for creating a digital survey;
encode the user-generated survey prompt into a vector space to generate a semantic label;
access a database comprising a plurality of survey graphs, wherein each survey graph represents a previously created and administered digital survey and comprises nodes corresponding to survey prompts;
identify, using connections of the plurality of survey graphs, a set of nodes with semantic labels corresponding to the semantic label of the user-generated survey prompt;
generate, based on response performance metrics associated with the set of nodes, a rephrased survey prompt comprising a rewording of the user-generated survey prompt; and
provide, for display on the administrator client device, the rephrased survey prompt.
18 . The system of claim 17 , further comprising instructions that, when executed by the at least one processor, cause the system to:
store semantic labels corresponding to historical survey prompts; and
present rewording suggestions based on the historical survey prompts.
19 . The system of claim 17 , further comprising instructions that, when executed by the at least one processor, cause the system to:
based on a node satisfying a survey prompt response metric, extract survey prompt wording corresponding to the node; and
provide the survey prompt wording from the node to the administrator client device as the rephrased survey prompt along with an option to use the rephrased survey prompt.
20 . The system of claim 17 , further comprising instructions that, when executed by the at least one processor, cause the system to:
traverse a first survey graph of the plurality of survey graphs to locate a node with a semantic label corresponding to the semantic label of the user-generated survey prompt; and
select the rephrased survey prompt based on a comparison of a response performance metric associated with the node with the response performance metrics associated with the set of nodes.