Author: James Allen, Nathanael Chambers, George Ferguson, Lucian Galescu, Hyuckchul Jung, Mary Swift, William Tayson |
Summary |
| This paper describes the system PLOW. The goal of the PLOW (Procedure Learning on the Web) is to build a system with which a user can teach the computer to perform tasks on the web. PLOW learns from both explicit demonstration of the task together with natural language instruction. The natural language play by play provides key information that allows rapid and robust learning of complex procedures including conditionals and iteration in one short session. PLOW demonstrates the power of an integrated approach to learning, combining deep natural language understanding, reasoning and machine learning. |
| |
| Strength |
| > Integrating natural language recognition and understanding (TRIPS, 2001) |
| |
“Play by play” mode, great user experience |
| Easier to identify parameters, boundaries of loops, termination conditions, build hierarchical structure, realize goals |
| > Generalization from one short task |
| |
Learn not only the task, but also the rule |
| > Error correction from users |
| |
“This is wrong. Here is another title” |
| > PLOW will confirm correctness from users when generating data from lists |
| > Less domain knowledge required, less training |
| > Close to "one-click automation" |
|
| |
| Weakness |
| > Some remarks of Evaluation: |
| |
PLOW was ensured of being able to learn 17 pre-determined test questions, other systems? |
| 10 new tasks have different levels of difficulties |
| No detailed analysis of evaluation result, so does PLOW really learn robust task models from a single example? Or just better on certain types of tasks |
| > Learning and reasoning relied on NL understanding: |
| |
encounters new concepts? |
| |
require certain patterns of speaking? Enough NL understanding capabilities? |
| > Still need one full work day to teach 3 simple tasks/person |
| > Users have to construct good task models, no error detection mechanism for users in PLOW |
|
| |
| Presentation: |
| |
download slides here |
| |
| Related works: |
| |
Sheepdog, 2004 (Sheep dog: Learning Procedures for Technical Support) |
| |
Tailor, 2005 (Task Learning by Instruction in Tailor) |
| |
CHINLE, 2008 (Recovering from Errors during Programming by Demonstration) |
| |
|