llm4s-core/org.llm4s/org.llm4s.knowledgegraph/org.llm4s.knowledgegraph.query/GraphQAPipeline

GraphQAPipeline

org.llm4s.knowledgegraph.query.GraphQAPipeline

class GraphQAPipeline(llmClient: LLMClient, graphStore: GraphStore, config: GraphQAConfig)

Graph-guided question answering pipeline.

Uses graph structure to provide rich context for LLM answers. The pipeline follows the pattern: identify entities -> traverse for context -> rank entities -> generate answer.

Multi-hop traversal follows relationship chains to gather evidence beyond direct neighbors. Citations track which nodes and edges contributed to the answer.

Value parameters

config: Pipeline configuration
graphStore: The graph store containing the knowledge graph
llmClient: The LLM client for entity identification and answer generation

Attributes

Example

val pipeline = new GraphQAPipeline(llmClient, graphStore)
val result: Result[GraphQAResult] = pipeline.ask("Who does Alice work with?")
result.foreach { r =>
 println(r.answer)
 r.citations.foreach(c => println(s"Source: ${c.nodeLabel} ${c.nodeId}"))
}

Graph

Supertypes

class Object

trait Matchable

class Any

Members list

Value members

Concrete methods

Answers a natural language question using graph-guided context.

Pipeline steps:

Identify entities mentioned in the question
Translate question to graph query
Execute query to gather context subgraph
Rank entities by importance (if enabled)
Generate answer with graph context and citations

Value parameters

question: The natural language question

Attributes

Returns: Right(result) with answer and citations, Left(error) on failure

Answers a question with an explicit pre-built graph query (skipping NL translation).

Useful when the caller already knows the appropriate query structure.

Value parameters

query: The pre-built graph query to execute
question: The natural language question for the answer generation step

Attributes

Returns: Right(result) with answer and citations, Left(error) on failure

In this article

Generated with