Natural language reasoning method and system
A technology of natural language and reasoning methods, applied in the field of natural language reasoning methods and systems, can solve problems such as unsupervised, poor model interpretability, no parameters, etc., to achieve the effect of improving expression ability, promoting interaction, and enhancing performance
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0059] The present invention provides a natural language reasoning method, such asfigure 1 shown, including:
[0060] S1 acquires and processes natural language sentence pairs that require language reasoning;
[0061] s2 uses a pre-trained algorithm model to predict the relationship between the processed sentence pairs;
[0062] Wherein, the algorithm model introduces an intra-sentence attention module and an inter-sentence attention module in the natural language reasoning model; the sentence pair mentioned in the present invention includes a "premise" sentence and a "hypothetical" sentence; the relationship of the sentence pair includes: Contradiction, implication, irrelevance;
[0063] For the training of the algorithm model in step S2, follow the steps below, specifically as figure 2 Shown:
[0064] Step 1. Select existing natural language sentence pairs and preprocess the data;
[0065] Specifically, natural language sentence pairs are selected from the natural langu...
Embodiment 2
[0105] In order to realize the above method, the present invention also provides a natural language reasoning system, such as Figure 5 shown, including:
[0106] Select the candidate sentence pair module to obtain and process the natural language sentence pairs that require language reasoning;
[0107] Loading a reasoning module, using a pre-trained algorithm model, to perform relationship prediction on the processed sentence pairs;
[0108] Wherein, the algorithm model introduces an intra-sentence attention module and an inter-sentence attention module into the natural language reasoning model; the relationship between sentence pairs includes: contradiction, implication, and irrelevance.
[0109] The loading reasoning module includes: model calculation module and model construction and training module;
[0110] Among them, the model calculation module is used to calculate the relationship prediction by using the model trained in the model construction and training module b...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More