String similarity

As explained in Knowledge Graphs and LLMs in Action

String similarity is a measure of how closely two textual character sequences match.

String similarity measures how closely textual sequences match. In entity resolution, it provides an initial signal for linking mentions that may refer to the same real-world entity. section 6.1.4

Use similarity as one signal in a graph-based resolution pipeline:

• Compare textual representations and create META_SIMILAR links for identical or very similar mentions. section 6.1.4
• Combine similarity with graph context, such as shared WORKS_FOR relations or research topics. section 6.1.4
• Set thresholds with domain rules: combine surname and first-name information rather than relying on surnames alone. section 6.1.4

Resolving person-name mentions

“Eleanor Smith” and “E. Smith” become stronger candidates when both mentions work for the same university; generic words such as “Foundation” should not create similarity links by themselves. section 6.1.4

Figure 6.3 Graph-based entity resolution example

Use the graph-based example to inspect how textual similarity and neighboring relations support entity resolution. section 6.1.4

How Knowledge Graphs and LLMs in Action covers it

How deep it goes

Defines a baseline using similarity links, domain thresholds, and contextual signals without detailing a full resolution system.

Examples

Uses abbreviated person names, shared universities, research topics, surnames, and organization names containing “Foundation.”

What it assumes

Assumes entity mentions, graph relations, and domain naming conventions; teaches their combination from the initial similarity baseline.

What you can do after

Apply it via graph links and contextual rules to prioritize possible matches while reducing false positives.

How it connects

Connects entity resolution with entity disambiguation, knowledge graphs, graph structure, and domain knowledge.

liveBookmanning.com

Read Knowledge Graphs and LLMs in Action on liveBook Get the book at manning.com

Also explained in

• 

  liveBookmanning.com
  Ensemble Methods for Machine LearningA string similarity metric is a function that assigns a numeric score indicating how similar two strings are.

Compare how 2 books explain string similarity →

Related concepts

• compressionmeasures · Compression is the transformation of information into a more compact representation to reduce storage or transmission requirements while retaining information needed for recovery or use.
• embeddinguses · A numerical vector representation of an item or input that encodes features or relationships useful for computational tasks.
• edit distancemeasured by · Edit distance is the minimum number of permitted operations needed to transform one string or sequence into another, commonly counting insertions, deletions, and substitutions.
• character-based string similarityincludes · Character-based string similarity compares strings according to character-level operations such as insertion, deletion, and substitution.

This explainer was generated by Manning's knowledge graph from the text of Knowledge Graphs and LLMs in Action. The section links point to the parts of the book it draws on.