liveBook · Manning

Fluency

As explained in Generative AI in Action

Language fluency is the grammaticality, naturalness, and readability of generated language.

Fluency measures the grammaticality, naturalness, and readability of generated language. It is evaluated alongside accuracy, coherence, and relevance to judge whether outputs suit real-world applications. section 12.1 section 12.6

Evaluate fluency as part of a broader, application-specific quality process:

  • Test outputs on their intended tasks, such as summarization, question answering, or translation. section 12.1
  • Combine automated metrics, manual review, user feedback, fair prompts, and ethical checks. section 12.1 section 9.3.2
  • Compare results with baselines and monitor production performance to identify improvements or remaining weaknesses. section 9.3.2

Scoring summary fluency with G-Eval

A G-Eval workflow uses Azure OpenAI GPT-4 to score article summaries for relevance, coherence, consistency, and fluency. In the reported comparison, the LLM summary scored 3 for fluency and the NLP summary scored 2; the workflow stores metric results for comparison. section 12.3.2

Listing 12.2 LLM-based evaluation metrics
EVALUATION_PROMPT_TEMPLATE = """                       #1
You will be given one summary written for an article. Your task [
is to rate the summary using one metric. Make sure you read 
and understand these instructions very carefully. 

Evaluation Criteria:
{criteria}

Evaluation Steps:
{steps}


Example:
Source Text:

{document}

Summary:
{summary}

Evaluation Form (scores ONLY):
- {metric_name}
"""

# Metric 1: Relevance             #2
RELEVANCY_SCORE_CRITERIA = """
Relevance(1-5) - selection of important content from the source. \
The summary should include only important information from the 
source document. \
Annotators were instructed to penalize summaries which contained 
redundancies and excess information.
"""

RELEVANCY_SCORE_STEPS = """        #3
1. Read the summary and the source document carefully.
2. Compare the summary to the source document and identify the 
   main points of the article.
3. Assess how well the summary covers the main points of the article,
   and how much irrelevant or redundant information it contains.
4. Assign a relevance score from 1 to 5.
"""

# Metric 2: Coherence             #4
COHERENCE_SCORE_CRITERIA = """
Coherence(1-5) - the collective quality of all sentences. \
...

COHERENCE_SCORE_STEPS = """                  #5
1. Read the article carefully and identify the main topic and key points.
2. Read the summary and compare it to the article. Check if the summary 
...

# Metric 3: Consistency                  #6
CONSISTENCY_SCORE_CRITERIA = """
Consistency(1-5) - the factual alignment between the summary and the 
summarized source.
...

CONSISTENCY_SCORE_STEPS = """               #7
1. Read the article carefully and identify its main facts and details.
2. Read the summary and compare it to the article. Check if the 
   summary ...

# Metric 4: Fluency                         #8
FLUENCY_SCORE_CRITERIA = """
Fluency(1-3): the quality of the summary in terms of grammar, spelling, punctuation, word choice, and sentence structure.
...

FLUENCY_SCORE_STEPS = """                 #9
Read the summary and evaluate its fluency based on the given criteria. 
Assign a fluency score from 1 to 3.
...

Inspect the evaluation prompt template and metric criteria used to turn fluency into a repeatable summary-scoring task. section 12.3.2

How Generative AI in Action covers it

How deep it goes
Defines fluency as a quality criterion and demonstrates scoring, but does not detail linguistic subcomponents or model internals.
Examples
Scores two article summaries with G-Eval using Azure OpenAI GPT-4 and compares fluency results alongside three other criteria.
What it assumes
Assumes familiarity with LLM outputs and evaluation metrics while applying the process to concrete NLP tasks and business use cases.
What you can do after
Evaluate generated summaries with G-Eval and combine scores with benchmarking, review, feedback, and production monitoring.
How it connects
Connects fluency with coherence, consistency, relevance, accuracy, benchmarking, and human evaluation.

Related concepts

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