Design Patterns for Explainable Agents (XAg). Rodriguez, S., Thangarajah, J., & Davey, A. In Proceedings of the 2024 International Conference on Autonomous Agents and Multiagent Systems, of AAMAS '24, pages 1621–1629, Auckland, New Zeland, 2024. ![Design Patterns for Explainable Agents (XAg) [pdf]](https://bibbase.org/img/filetypes/pdf.svg "pdf")
Paper abstract bibtex The ability to explain the behaviour of the AI systems is a key aspect of building trust, especially for autonomous agent systems - how does one trust an agent whose behaviour can not be explained? In this work, we advocate the use of design patterns for developing explainable-by-design agents (XAg), to ensure explainability is an integral feature of agent systems rather than an “add-on” feature. We present TriQPAN (Trigger, Query, Process, Action and Notify), a design pattern for XAg. TriQPAN can be used to explain behaviours of any agent architecture and we show how this can be done to explain decisions such as why the agent chose to pursue a particular goal, why or why didn’t the agent choose a particular plan to achieve a goal, and so on. We term these queries as direct queries. Our framework also supports temporal correlation queries such as asking a search and rescue drone, “which locations did you visit and why? ”. We implemented TriQPAN in the SARL agent language, built-in to the goal reasoning engine, affording developers XAg with minimal overhead. The implementation will be made available for public use. We describe that implementation and apply it to two case studies illustrating the explanations produced, in practice.
@inproceedings{Rodriguez2024Triqpan,
address = {Auckland, New Zeland},
series = {{AAMAS} '24},
title = {Design {Patterns} for {Explainable} {Agents} ({XAg})},
url = {https://www.ifaamas.org/Proceedings/aamas2024/pdfs/p1621.pdf},
abstract = {The ability to explain the behaviour of the AI systems is a key aspect
of building trust, especially for autonomous agent systems - how
does one trust an agent whose behaviour can not be explained? In
this work, we advocate the use of design patterns for developing
explainable-by-design agents (XAg), to ensure explainability is an
integral feature of agent systems rather than an “add-on” feature.
We present TriQPAN (Trigger, Query, Process, Action and Notify), a
design pattern for XAg. TriQPAN can be used to explain behaviours
of any agent architecture and we show how this can be done to
explain decisions such as why the agent chose to pursue a particular
goal, why or why didn’t the agent choose a particular plan to
achieve a goal, and so on. We term these queries as direct queries.
Our framework also supports temporal correlation queries such as
asking a search and rescue drone, “which locations did you visit
and why? ”. We implemented TriQPAN in the SARL agent language,
built-in to the goal reasoning engine, affording developers XAg
with minimal overhead. The implementation will be made available
for public use. We describe that implementation and apply it to two
case studies illustrating the explanations produced, in practice.},
booktitle = {Proceedings of the 2024 {International} {Conference} on {Autonomous} {Agents} and {Multiagent} {Systems}},
author = {Rodriguez, Sebastian and Thangarajah, John and Davey, Andrew},
year = {2024},
pages = {1621--1629},
}
Downloads: 0
{"_id":"yWF7nmxy5AMsXwwN7","bibbaseid":"rodriguez-thangarajah-davey-designpatternsforexplainableagentsxag-2024","author_short":["Rodriguez, S.","Thangarajah, J.","Davey, A."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","address":"Auckland, New Zeland","series":"AAMAS '24","title":"Design Patterns for Explainable Agents (XAg)","url":"https://www.ifaamas.org/Proceedings/aamas2024/pdfs/p1621.pdf","abstract":"The ability to explain the behaviour of the AI systems is a key aspect of building trust, especially for autonomous agent systems - how does one trust an agent whose behaviour can not be explained? In this work, we advocate the use of design patterns for developing explainable-by-design agents (XAg), to ensure explainability is an integral feature of agent systems rather than an “add-on” feature. We present TriQPAN (Trigger, Query, Process, Action and Notify), a design pattern for XAg. TriQPAN can be used to explain behaviours of any agent architecture and we show how this can be done to explain decisions such as why the agent chose to pursue a particular goal, why or why didn’t the agent choose a particular plan to achieve a goal, and so on. We term these queries as direct queries. Our framework also supports temporal correlation queries such as asking a search and rescue drone, “which locations did you visit and why? ”. We implemented TriQPAN in the SARL agent language, built-in to the goal reasoning engine, affording developers XAg with minimal overhead. The implementation will be made available for public use. We describe that implementation and apply it to two case studies illustrating the explanations produced, in practice.","booktitle":"Proceedings of the 2024 International Conference on Autonomous Agents and Multiagent Systems","author":[{"propositions":[],"lastnames":["Rodriguez"],"firstnames":["Sebastian"],"suffixes":[]},{"propositions":[],"lastnames":["Thangarajah"],"firstnames":["John"],"suffixes":[]},{"propositions":[],"lastnames":["Davey"],"firstnames":["Andrew"],"suffixes":[]}],"year":"2024","pages":"1621–1629","bibtex":"@inproceedings{Rodriguez2024Triqpan,\n\taddress = {Auckland, New Zeland},\n\tseries = {{AAMAS} '24},\n\ttitle = {Design {Patterns} for {Explainable} {Agents} ({XAg})},\n\turl = {https://www.ifaamas.org/Proceedings/aamas2024/pdfs/p1621.pdf},\n\tabstract = {The ability to explain the behaviour of the AI systems is a key aspect\nof building trust, especially for autonomous agent systems - how\ndoes one trust an agent whose behaviour can not be explained? In\nthis work, we advocate the use of design patterns for developing\nexplainable-by-design agents (XAg), to ensure explainability is an\nintegral feature of agent systems rather than an “add-on” feature.\nWe present TriQPAN (Trigger, Query, Process, Action and Notify), a\ndesign pattern for XAg. TriQPAN can be used to explain behaviours\nof any agent architecture and we show how this can be done to\nexplain decisions such as why the agent chose to pursue a particular\ngoal, why or why didn’t the agent choose a particular plan to\nachieve a goal, and so on. We term these queries as direct queries.\nOur framework also supports temporal correlation queries such as\nasking a search and rescue drone, “which locations did you visit\nand why? ”. We implemented TriQPAN in the SARL agent language,\nbuilt-in to the goal reasoning engine, affording developers XAg\nwith minimal overhead. The implementation will be made available\nfor public use. We describe that implementation and apply it to two\ncase studies illustrating the explanations produced, in practice.},\n\tbooktitle = {Proceedings of the 2024 {International} {Conference} on {Autonomous} {Agents} and {Multiagent} {Systems}},\n\tauthor = {Rodriguez, Sebastian and Thangarajah, John and Davey, Andrew},\n\tyear = {2024},\n\tpages = {1621--1629},\n}\n\n","author_short":["Rodriguez, S.","Thangarajah, J.","Davey, A."],"key":"Rodriguez2024Triqpan","id":"Rodriguez2024Triqpan","bibbaseid":"rodriguez-thangarajah-davey-designpatternsforexplainableagentsxag-2024","role":"author","urls":{"Paper":"https://www.ifaamas.org/Proceedings/aamas2024/pdfs/p1621.pdf"},"metadata":{"authorlinks":{}}},"bibtype":"inproceedings","biburl":"https://api.zotero.org/users/629018/collections/Z3Y89R5J/items?key=7FwOYSA57XqwE4i5WOHIlvZZ&format=bibtex&limit=100","dataSources":["dpS26Nm5wBb9pZTfk","2xjJWWKzozqCYv3fA","yeCWMsXdeEPFrdJ8M"],"keywords":[],"search_terms":["design","patterns","explainable","agents","xag","rodriguez","thangarajah","davey"],"title":"Design Patterns for Explainable Agents (XAg)","year":2024}