Semantic Parsing for Question Answering over Knowledge Graphs

Sijia Wei,Wenwen Zhang,Qisong Li,Jiang Zhao

2024-01-28

Abstract:In this paper, we introduce a novel method with graph-to-segment mapping for question answering over knowledge graphs, which helps understanding question utterances. This method centers on semantic parsing, a key approach for interpreting these utterances. The challenges lie in comprehending implicit entities, relationships, and complex constraints like time, ordinality, and aggregation within questions, contextualized by the knowledge graph. Our framework employs a combination of rule-based and neural-based techniques to parse and construct highly accurate and comprehensive semantic segment sequences. These sequences form semantic query graphs, effectively representing question utterances. We approach question semantic parsing as a sequence generation task, utilizing an encoder-decoder neural network to transform natural language questions into semantic segments. Moreover, to enhance the parsing of implicit entities and relations, we incorporate a graph neural network that leverages the context of the knowledge graph to better understand question representations. Our experimental evaluations on two datasets demonstrate the effectiveness and superior performance of our model in semantic parsing for question answering.

Computation and Language

What problem does this paper attempt to address?

The paper primarily focuses on the research and improvement of semantic parsing techniques in Knowledge Graph Question Answering (KGQA). Specifically, the paper attempts to address the following key issues:

1. **Understanding complex question semantics**: This includes understanding implicit entities, relationships, and complex constraints in questions, such as time, ordinality, and aggregation operations.

2. **Improving the accuracy of semantic parsing**: By proposing a novel method, the paper aims to more accurately map natural language questions to the knowledge graph, thereby obtaining more precise answers.

3. **Handling multi-intent combination problems**: In domain-specific knowledge graph question answering, questions often contain multiple intents, each with certain constraints on the expected results. The paper attempts to address this challenge by splitting and combining these intents.

### Main Contributions of the Paper

1. **Proposed a semantic parsing framework for knowledge graph question answering based on semantic fragments** (SPEDN), defining a series of semantic fragment patterns to represent the semantics of questions.

2. **Modeled the semantic parsing of questions as the problem of generating a sequence of semantic fragments**, and used graph neural networks to parse the questions to generate the sequence of semantic fragments. This method was validated on two datasets and achieved good results.

3. **Addressed the problem of identifying implicit entities and relationships**: By combining the contextual information of the knowledge graph, a contextual dictionary of the question was constructed as the input to the graph neural network, thereby improving the effectiveness of semantic parsing.

### Summary

In summary, the paper aims to enhance the ability of knowledge graph question answering systems in question understanding and semantic parsing, particularly in handling questions with implicit information and multi-intent combinations, by proposing the SPEDN framework. This method effectively improves the accuracy and efficiency of question understanding by decomposing questions into a series of semantic fragments and using techniques such as graph neural networks for parsing.