Foundations of Artificial Intelligence (FAI) Group

Seminar: Explainable AI Planning

Basics. Seminar, 7 graded ECTS points.

The seminar will be run in a block format. There will be an initial meeting on Wednesday, October 16, 16:15--17:45. All student presentations will be given on a single day after the end of term. A detailed schedule is given below.

All meetings will take place in room 3.06, Building E1 1. The seminar language is English throughout.

Supervisors for the seminar are Prof. Jörg Hoffmann, and Daniel Höller.

Your task will be to read and understand a piece of research, to write a summary paper in your own words, to give a presentation, and to provide detailed feedback for the paper and presentation of a fellow student.

All email interaction must be pre-fixed with "[XAIP19-20]" in the email subject.

No plagiarism. It is Ok (and encouraged!) to use web resources to further your understanding of your assigned topic. However, it is inadmissible to use pieces of such material for your summary paper or presentation. Any plagiarism will result in disqualification from the seminar.

Content. Model-based approaches to AI are well suited to explainability in principle, given the explicit nature of their world knowledge and of the reasoning performed to take decisions. AI Planning in particular is relevant in this context as a generic approach to action-decision problems. Indeed, explainable AI Planning (XAIP) has received interest since more than a decade, and has been taking up speed recently along with the general trend to explainable AI. This seminar provides an overview of the area, covering its major work lines pertaining to contrastive explanation, model reconciliation, hierarchical planning with user interaction, and the explanation of unsolvability.

Prerequisites. Participants must have successfully completed an introductory course in Artificial Intelligence. They should be familiar with automatic planning at least to the extent of the material covered in the Artificial Intelligence course; successful participation in one of our AI Planning courses will be an advantage, but is not absolutely necessary to follow the seminar.

Registration. Is via the central seminar registration system.

Grading. The final grading will be based, in this order of importance, on:

• The quality of your final presentation.
• The quality of your final summary paper.
• The quality of the feedback you provide to your mentee student (see below).
• Your participation in the discussions during the block seminar.

Summary Paper. For the summary paper, you must use this tex template. You are required to read at least 2 related papers, for the related work section. You are allowed to modify the section structure given in the template if, for whatever reason, this is more adequate for the work you are summarizing.

The seminar paper should be about 4 pages long (not counting the literature list, and in the double-column format of the template). This is a rough guideline, not a strict rule. If you need, say, 5-6 pages to do your paper justice then definitely do so.

Schedule and Deadlines (tentative!).

• October 23, 16:15--17:45: Initial meeting.

• October 28: Send a ranked list of the topics you would like to take, by email to Jörg Hoffmann. That is, send something like "area 1.2, area 2.2, area 1.1, area 4.2, area 3.1". Please include into this list all topics that you would be willing to accept. The list must contain at least 5 topics.

  Note that each topic is associated with a mentee student (to whom you will provide feedback, see the following deadlines); and a mentor student (who will provide feedback to you, see the following deadlines). The mentee/mentor assignment will be a "cycle" through each of the four topic areas as listed below: within areas with 2 topics, the two students mentor each other; within areas with 3 topics, the student with topic 1 mentors the student with topic 2, who mentors the student with topic 3, who mentors the student with topic 1. If you want to team up with someone specific, please do state that in your email.

• October 30: Receive your topic.

  Read the material associated with your topic carefully, and prepare an initial version of your summary paper, using the tex template given above.

• End of November: Make an appointment with your supervisor to discuss your paper. You should give a brief oral summary of the paper (5--10 minutes). Take the opportunity to ask questions. For students supervised by Daniel Höller, the appointment will be via skype. For students supervised by Jörg Hoffmann, the appointment will be on November 22, with a 45-minute slot for each student. The time slots are: 1. 10:00--10:45, 2. 10:45--11:30, 3. 11:30--12:15, 4. 12:15--13:00, 5. 14:00--14:45, 6. 14:45--15:30, 7. 15:30--16:15. Please email Jörg Hoffmann at the latest by November 17 with your availability, i.e., with the subset of time slots that would work for you. That is, send something like "1,2,5,7".

  NOTE: The following deadlines marked with "(ca.)" are meant as a guideline. You are required to do these things, but if you do them 3-4 days earlier or later, that is no problem.

• December 2 (ca.): Send your summary paper to your mentor student (cc supervisors).
• December 16 (ca.): Send feedback regarding the summary paper to your mentor student (cc supervisors).
• January 10 (ca.): Send revised summary paper to your mentor student (cc supervisors).
• January 15 (ca.): Send feedback regarding the revised summary paper to your mentor student (cc supervisors).
• January 17 (ca.): Send presentation slides to mentor student (cc supervisors).
• January 24 (ca.): Send feedback regarding the presentation slides to your mentor student (cc supervisors).
• January 31 (ca.): Send revised presentation slides to mentor student (cc supervisors).
• February 05 (ca.): Send feedback regarding the revised presentation slides to your mentor student (cc supervisors).

• February 15: Send your final summary paper by email to Jörg Hoffmann.
• February 18: Give a presentation (20 minutes talk, plus 10 minutes discussion) in the block seminar. Attendance to all talks is required. Please try to stick to the 20 minutes time slot for your talk; it should not be a lot shorter, nor a lot longer.

Topics. Each participant will be assigned one topic, each of which consists of one paper. The overall amount and difficulty of the material associated with each topic is roughly balanced.

Area 1: Contrastive Explanation. Supervisor: Jörg Hoffmann

1. A framework for contrastive explanation. Paper: Area 1 Topic 1
2. Compilation-based question answering in contrastive explanation. Paper: Area 1 Topic 2
3. Explaining the space of plans through plan-property dependencies. Paper: Area 1 Topic 3

Area 2: Model Reconciliation. Supervisor: Jörg Hoffmann

1. XAIP as model reconciliation against a known user model. Paper: Area 2 Topic 1
2. XAIP as model reconciliation against an unknown user model. Paper: Area 2 Topic 2

Area 3: Explaining Unsolvability. Supervisor: Jörg Hoffmann

1. Excuses why no plan could be found. Paper: Area 3 Topic 1
2. Hierarchical abstractions. Paper: Area 3 Topic 2

Area 4: Explainability in Hierarchical Planning. Supervisor: Daniel Höller

1. Application of explanation and other techniques for user instructions in home-theater assembly. Paper: Area 4 Topic 1
2. Plan-step explanation in hierarchical planning. Paper: Area 4 Topic 2
3. User-friendly plan linearization. Paper: Area 4 Topic 3