NLG applications enhanced by knowledge base

Inorder to utilize the full capacity of Knowledge bases ,joint training of neural models on multiple tasks like question understanding ,fact retrieval from KB is done if the target task is answer sequence generation.

Design Supervised Tasks around KB for Joint Optimization

Prior distribution is used to approximate posterior distributions over knowledge base inorder to select better facts from the knowledge base for an input output pair.

Enhance Incorporation by Selecting KB or Facts in KB

Knowledge base has  large scale information where data is maintained in “factual triplets”. KBs enhance text generation models 

- Design Supervised Tasks around KB for Joint Optimization
- Enhance Incorporation by Selecting KB or Facts in KB


San Jose State University

 -  NLG Enhanced by Knowledge Base
 - NLG Enhanced by Knowledge Graph
 - NLG enhanced by Grounded Text

NLG ENHANCED BY EXTERNAL KNOWLEDGE

A Survey of Knowledge-Enhanced Text Generation

NLG Enhanced by Knowledge Base

Knowledge graph is a structured way of representing facts in the form of connections between entities ,relations and semantic descriptions. Different methods to study the relationship between KG and input/output sequences are

- Incorporate Knowledge Graph Embeddings into Language Generation
- Transfer Knowledge into Language Model with Knowledge Triplet Information.
- Perform Reasoning over Knowledge Graph via Path Finding Strategies.
- Improve the Graph Embeddings with Graph Neural Networks


NLG Enhanced by Knowledge Graph 

Ground text is the textual data that is not available in any structured form or in any database, these are available in  online resources like encyclopedias, social media , shopping websites.
knowledge grounded text is often taken as an important external knowledge source to help with a variety of NLG applications.

 - Guiding Generation with Retrieved Information
- Modeling Background Knowledge into Response Generation


Learn Before

Related

Learn After