Response generation for conversational computing interface

1. A computer-implemented method of responding to a conversational event, comprising:

enacting, by a conversational computing interface, an initial computer-executable plan based on a conversational event received by the conversational computing interface, wherein the initial computer-executable plan is configured to output an initial value based on the conversational event;

selecting, by the conversational computing interface based on the initial value, a description template that requires additional information not output by the initial computer-executable plan;

selecting, by the conversational computing interface based on the initial value, an extended computer-executable plan configured to output the additional information;

enacting, by the conversational computing interface, the extended computer-executable plan to fill in the description template with the additional information;

outputting a plurality of candidate responses to the conversational event based on a plurality of different description templates, each different description template filled with additional information from a corresponding extended computer-executable plan; and

outputting, via the conversational computing interface, a response to the conversational event based on one of the different description templates filled in with additional information.

2. The computer-implemented method of claim 1 , wherein the initial value is a result of an application programming interface (API) call issued by the initial computer-executable plan, the additional information includes an answer to a further API call, and the extended computer-executable plan is configured to issue the further API call to retrieve the answer to the further API call.

3. The computer-implemented method of claim 1 , further comprising recursively enacting further-extended computer-executable plans to recursively output further-additional information and further fill the description template with the further-additional information.

4. The computer-implemented method of claim 1 , wherein selecting the description template and selecting the extended computer-executable plan are based on selecting an applicable generation rule of a plurality of generation rules, wherein the applicable generation rule defines the description template and the extended computer-executable plan.

5. The computer-implemented method of claim 4 , wherein selecting the generation rule of the plurality of generation rules includes recognizing a syntactic symbol indicating a syntactic role of the response to the conversational event, and filtering the plurality of generation rules to select the applicable generation rule from a subset of generation rules having a matching syntactic symbol.

6. The computer-implemented method of claim 4 , wherein selecting the applicable generation rule of the plurality of generation rules includes assessing, for each candidate generation rule of the plurality of generation rules, a correspondence between the initial value output by the initial computer-executable plan and a plan matcher of the candidate generation rule.

7. The computer-implemented method of claim 6 , wherein the plan matcher is configured to match any value that satisfies a computer-executable predicate function defined for values of a predefined data type.

8. The computer-implemented method of claim 6 , wherein the plan matcher is configured to match one or more structural features of the initial computer-executable plan.

9. The computer-implemented method of claim 6 , wherein the plan matcher is configured to match a specific value.

10. The computer-implemented method of claim 1 , further comprising assessing a score for each different candidate response and outputting the response having a highest score of the plurality of candidate responses.

11. The computer-implemented method of claim 1 , further comprising assessing, for each different candidate response, a score of a description template associated with the response and outputting the response having a highest aggregate score of one or more associated description templates.

12. A computer-implemented method of responding to a conversational event, comprising:

enacting, by a conversational computing interface, an initial computer-executable plan based on a conversational event received by the conversational computing interface, wherein the initial computer-executable plan is configured to output an initial value based on the conversational event;

selecting, by the conversational computing interface, an extended computer-executable plan based on determining that the initial value is insufficient for generating an extended description responsive to the conversational event;

enacting, by the conversational computing interface, the extended computer-executable plan to output additional information beyond what the initial computer-executable plan is configured to output, the additional information sufficient for generating the extended description responsive to the conversational event;

outputting a plurality of candidate responses to the conversational event based on filling in a plurality of different description templates, wherein each different description template is filled in with information from a corresponding extended computer-executable plan; and

outputting, via the conversational computing interface, a response to the conversational event including the extended description based on one of the different description templates filled in with the additional information.

13. The computer-implemented method of claim 12 , further comprising selecting, based on the initial value, a description template that requires the additional information, wherein the response to the conversational event is based on the description template filled in with the additional information.

14. The computer-implemented method of claim 12 , wherein the initial value is a result of an application programming interface (API) call issued by the initial computer-executable plan, the additional information includes an answer to a further API call, and the extended computer-executable plan is configured to issue the further API call to retrieve the answer to the further API call.

15. The computer-implemented method of claim 12 , further comprising:

filling a description template with the additional information;

recursively enacting further-extended computer-executable plans to recursively output further-additional information; and

filling the description template with the further-additional information.

16. The computer-implemented method of claim 12 , wherein selecting the extended computer-executable plan includes:

recognizing a syntactic symbol indicating a syntactic role of the response to the conversational event; and

filtering a plurality of generation rules to select an applicable generation rule from a subset of generation rules having a matching syntactic symbol, wherein the applicable generation rule defines the extended computer-executable plan.

17. The computer-implemented method of claim 12 , wherein selecting the extended computer-executable plan includes assessing, for each candidate generation rule of a plurality of generation rules, a correspondence between the initial value and a plan matcher of the candidate generation rule, wherein an applicable generation rule defines the extended computer-executable plan.

18. A conversational computing interface system, comprising:

a logic device; and

a storage device holding instructions executable by the logic device to:

instantiate a conversational computing interface;

enact, by the conversational computing interface, an initial computer-executable plan based on a conversational event received by the conversational computing interface, wherein the initial computer-executable plan includes computer-executable bytecode in a data-flow programming language, and wherein the initial computer-executable plan is configured to output an initial value based on the conversational event;

select, by the conversational computing interface, an extended computer-executable plan including additional data-flow program fragments based on determining that the initial value is insufficient for generating an extended description responsive to the conversational event;

enact, by the conversational computing interface, the extended computer-executable plan to output additional information beyond what the initial computer-executable plan is configured to output, the additional information sufficient for generating the extended description responsive to the conversational event; and

output, via the conversational computing interface, a response to the conversational event including the extended description.