Difference between revisions of "Vulcan/SystemTarget"
From Knowitall
(→Method) |
(→Method) |
||
Line 24: | Line 24: | ||
<pre> | <pre> | ||
− | 1. Hand generate the set of rules and axioms in a human-readable format (click [[|here]] for an example. Convert this into the Tuffy MLN format. | + | 1. Hand generate the set of rules and axioms in a human-readable format (click [[|here]] for an example. |
− | 2. Run Open IE 4.0 on each input sentence. Extract the best tuple: One that covers the sentence best. If there is a tie use the tuple with the longest relation phrase. | + | Convert this into the Tuffy MLN format. |
+ | 2. Run Open IE 4.0 on each input sentence. Extract the best tuple: One that covers the sentence best. | ||
+ | If there is a tie use the tuple with the longest relation phrase. | ||
3. | 3. | ||
</pre> | </pre> | ||
== System Components == | == System Components == |
Revision as of 03:53, 22 August 2013
This page describes the features in the first implementation of the system.
I/O
- Input
- Natural language sentences as input.
- The system will handle propositions that correspond to three questions.
- X is the best conductor of electricity. X = {Iron nail, wax crayon, plastic cup, rubber boat}
- X causes leaves of a plant to become larger. X = {Growth, A repair, Germination, Decomposition}
- X helps a fox find food. X = {A sense of smell, ...}
- Output
- Marginal inference probability for each proposition according to Tuffy.
- Set of rules and axioms used (human-readable and MLN format). Note this is going to be hand generated in this iteration.
- Debugging output that shows the set of axioms, rules and new facts used in the inference.
Method
1. Hand generate the set of rules and axioms in a human-readable format (click [[|here]] for an example. Convert this into the Tuffy MLN format. 2. Run Open IE 4.0 on each input sentence. Extract the best tuple: One that covers the sentence best. If there is a tie use the tuple with the longest relation phrase. 3.