CS298 Proposal
Online Collaborative Time Management System using Artificial Intelligence
Anand Sivaramakrishnan (anandsk123@gmail.com)
Advisor: Dr. Chris Pollett
Committee Members: Dr. Robert Chun, Dr. Teng Moh
Abstract:
This project will implement an intelligent online collaborative system for scheduling different people's TO DO lists according to their daily routine. TO DO lists differ
from calendar events in that the time frame for the event is less specified. A task might have a due date, and other tasks called prerequisites which need to be completed first.
We will design a GUI interface to make it easy to add prerequisites (if any) for TO DO items. It will support different people adding items to other peoples' TO DO lists in much the same way
as Google calendar supports adding new events to other peoples' calendars. One way to organize the order to do events using artificial intelligence techniques is to create a partial or total order plan.
Our system will attempt to do this on people's TO DO lists. What is novel, however, in our setting is that it is online, that is, we do not know all of the events a priori as events can be added
dynamically.
CS297 Results
- Implemented the Total Order Planning Algorithm and tried it on some test cases(situations).
- Developed a Partial Order Planner and tried this also on some test cases(situations), mostly text book examples.
- Converted the Partial Order Planner of Deliverable 2 in an Object Oriented form.
- Sketched web GUIs for adding prerequistes and actions to TO DO events.
Proposed Schedule
| Week 1:
Aug 26 to Sep 2 | Make the CS 298 proposal. |
| Week 2 - Week 4:
Sep 2 to Sep 23 | Design a database schema and structure it such that it will be robust and handle all kinds of user generated
inputs. |
| Week 5 - Week 7:
Sep 23 - Oct 14 | Modify the Partial Order Planning algorithm and make it work for real World Scenarios. Also, connect it
to the above database |
| Week 8 - Week 11:
Oct 14 - Nov 11 | Design and code the User Interface and connected to the earlier deliverables. Also draft report
and submit to Graduate Office |
| Week 12 - Week 14:
Nov 11 - Dec 2 | Testing the above system and also make changes to the CS 298 Report that were suggested by the Graduate Office. |
| Week 15:
Dec 2 - Dec 9 | Defense in front of the committee. |
Key Deliverables:
- Software
Deliverable_1: Storing the user details as well as the user generated actions, preconditions and plans, and collaborating
data between different tables in a well structured database will be the first deliverable. This project is something that will be launched on the web.
It will be dynamic in nature. This means it will take the user generated input and accordingly formulate the output. For this to be implemented,
I have to structure a database which can make this dynamic nature practical.
Deliverable_2: My second deliverable will be to ensure that the algorithm works for all kinds of complex real world scenarios between
collaborative groups. The Partial Order Planing Algorithm already implemented is an algorithm which may have to be modified to work for the kind of
system I am building. This algorithm has never been used for examples other than that of the text book and a few other easy ones.
Deliverable_3: Making a very user friendly and easy interface will be the third deliverable. Also, extensive testing of the UI.
This will be very important because if the user interface is not easy, then the site will not attract potential users. Connecting this UI to
the modified back end partial order algorithm and the newly structured database will be the done.
- Report
Final report consisting the detailed description of software used, algorithms implemented, design patterns used in the project will be
delivered. The report will also contain the prior work done in this field, and how this project is innovative and different than the work that
is already done. It will also contain the experiements that will be already done on a few users, and their experience with the system.
Innovations and Challenges
- This online planner as we all know will be used by many people and will be appreciated or criticized by all of them. Therefore,
it has to be ensured that this system is robust in terms of producing logical results in majority of the cases. For this to happen, a good database
structure and schema is very important, this will be accomplished in the first deliverable. The inputs from users is going to be constant
and dynamic, and will be used for making their plans.
- One of the challenges in this project is that of using an existing Partial Order Planning algorithm and modifying it such that it is
robust for all kinds of inputs from users in various situations. This will be done in the second deliverable. This algorithm has been
designed and tested for text book examples. Therefore modifying this algorithm such that it will give logical solutions to collaborative
groups within specific time deadlines is a challenge.
- The algorithm and the database structure will be both new in structure and schema. Therefore making a UI that will be comfortable to the
user and to be connected to this complex backend will be one of the challenges in this project. The third deliverable involves this completion.
References:
Rajeev Motwani, Prabhakar Raghavan. (1995). Randomized Algorithms. Cambridge University Press.
Peter Norvig, Stuart Russell. (1995) Artificial Intelligence: A Modern Approach. Prentice Hall Series.
Sussanne Biundo, Maria Fox. (1999). Recent Advances in AI Planning. ECP, Springer.
Craig Knoblock, Qiang Yang. (1997). Relating the Performance of Partial Order Planning Algorithms to Domain Features . SIGART Bulletin, Vol. 6, No. 1, 8-15
Qiang Yang, M Pollack. (1997) Intelligent Planning: A Decomposition and Abstraction Based Approach (Artificial Intelligence).
|