Graph Based Semi-Supervised Learning for Entailment Ranking | So similarity between two q/a pairs 50,-, 533-, is represented with wij E 32””, i.e., edge weights , and is measured as: |
Graph Based Semi-Supervised Learning for Entailment Ranking | As total entailment scores get closer, the larger their edge weights would be. |
Graph Based Semi-Supervised Learning for Entailment Ranking | Thus, we modify edge weights in (1) as follows: |
Graph Summarization | Our idea of summarization is to create representative vertices of data points that are very close to each other in terms of edge weights . |
Graph Summarization | We identify the edge weights wfj between each node in the boundary Bf via (1), thus the boundary is connected. |
Graph Summarization | If any testing vector has an edge between a labeled vector, then with the usage of the local density constraints, the edge weights will not not only be affected by that labeled node, but also how dense that node is within that part of the graph. |
The Metric-based Framework | Formally, it is a function d :C><C a [Rh where C is the set of terms in T. An ontology metric d on a taxonomy T with edge weights w |
The Metric-based Framework | for any term pair (ox,cy)EC is the sum of all edge weights along the shortest path between the pair: |
The Metric-based Framework | In the training data, an ontology metric d(c,,,cy) for a term pair (obey) is generated by assuming every edge weight as 1 and summing up all the edge weights along the shortest path from C, to Cy. |