Event causality extraction method based on derivative cue learning
A causal relationship and event technology, applied in the field of information extraction, can solve the problems of difficult to extract models, difficult to label cost estimation, high labeling cost, etc., to achieve improved ability, increase learning efficiency and robustness, and make up for the number of positive samples insufficient effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0035] Example 1: see figure 1 , an event causal relationship extraction method based on derivative cue learning, the specific steps are as follows:
[0036] Step 1) Construct two derivative tasks of causality extraction;
[0037] Step 2) Build a prompt template for the causal relationship extraction task;
[0038] Step 3) Build a prompt template for derived tasks;
[0039] Step 4) Construct a derived cue causality extraction model with gated units;
[0040] Step 5) Train the causal relationship extraction model through supervised learning based on the teacher mechanism;
[0041] Step 6) Cue-based causality extraction.
[0042] Among them, in step 1), a derivative task of causal relationship extraction needs to be constructed.
[0043] First, based on the need to predict causal cue words related to event pairs in a sentence to show causal relationship, we constructed a derivative task of causal cue word prediction. The input of causal cue word prediction is a sentence and...
Embodiment 2
[0053] Example 2: see figure 1 , the input text of event causality extraction is defined as , the event pair is , represents the source event, Represents the target event, both a specific trigger word (i.e. sequence The symbol in ), an event causal relationship extraction based on derivative cue learning includes the following steps: step 1) constructing a derivative task of causal relationship extraction;
[0054] First, based on the need to predict the causal cue words related to the event pair in the sentence to show the causal relationship, we constructed a derivative task of causal cue word prediction. The causal cue word prediction is derived from the event causal relationship extraction task, and its input is a sentence and the event pair in the sentence , whose goal is to predict the cue words in the sentence that determine the causal relationship between the two events. If the two events are not causally related or there is no obvious cue word in the ...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More