Skip to content

Ehud Reiter's Blog

Ehud's thoughts and observations about Natural Language Generation

  • Home
  • Blog Index
  • About
  • What is NLG
  • Publications
  • Resources
  • University
  • Contact

Author: ehudreiter

building NLG systems

We need to understand what users want!

Aug 8, 2022 ehudreiterLeave a comment

We can build much better NLG systems if we understand what users want the systems to do! This may sound trite, but there is very little research in the academic community in understanding user needs and requirements, which is a shame and indeed lost opportunity.

academics

An Architecture for Data-to-Text Systems

Jul 22, 2022Jul 27, 2022 ehudreiterLeave a comment

I was very happy to win an INLG Test of Time award for my paper “An Architecture for Data-to-Text Systems”, so I thought I’d write a few comments on it.

personal

Cycling through Southwest England

Jul 4, 2022Aug 5, 2022 ehudreiterLeave a comment

A travelogue about a recent bike trip. After two years of being limited in my holidays by Covid, it was great to finally be able to do some cycle touring again!

academics

Real-World Impact of Academic Research

Jun 16, 2022 ehudreiterLeave a comment

Society (and most funding agencies) want to see real-world benefits or “impact” from academic research. Of course not all research will have real-world impact, and impact may take years or decades to appear! I share some thoughts on types of impact, barriers to impact, and my personal experiences.

evaluation

Lets use error annotations to evaluate systems!

Jun 1, 2022 ehudreiterLeave a comment

I am excited by the idea of using error annotation to evaluate NLG systems, where domain experts or other knowledgeable people mark up individual errors in generated texts. I think this is usually more meaningful and gives better insights that asking crowdworkers to rate or rank texts, which is how most human evaluations are currently done.

building NLG systems

NLG=Task+Data+Model/Alg+Eval

May 20, 2022May 20, 2022 ehudreiterLeave a comment

Progress in NLG requires understanding what users want, creating high quality data sets, building models and algorithms, and thoroughly evaluating systems. I remain disappointed that the research community seems fixated on building models and pays much less attention to user needs, datasets, and evaluation.

evaluation

We need more extrinsic (task) evaluation!

May 9, 2022 ehudreiter1 Comment

The most meaningful evaluation is when we test whether an NLG system actually achieves its communicative goal, eg helps people make better decisions or write documents faster. Unfortunately such “extrinsic” or “task” evaluation is rare in NLP in 2002, we need to see more such evaluations!

academics

Can I present my paper twice?

Apr 26, 2022 ehudreiterLeave a comment

I’ve come to realise that there is some confusion, especially amongst newcomers to NLP/AI, about when a research paper can be presented at two venues. I try to explain the rules and principles as I understand them.

evaluation

Why is ROUGE so popular?

Apr 10, 2022Apr 13, 2022 ehudreiter2 Comments

The ROUGE metric dominates evaluation of summarisation, and I do not understand why. I am not aware of good evidence that ROUGE predicts utility, and recent work by one of my students shows that character-level edit (Levenshtein) distance against a reference text is a better predictor of utility than ROUGE.

evaluation

Humans make mistakes too

Apr 3, 2022Apr 4, 2022 ehudreiter1 Comment

Some of my PhD students have recently looked at how many mistakes people (professionals, not Turkers) make when they do NLG-like tasks. The number of mistakes is considerably higher than we expected (although still much lower than the number of mistakes made by current neural NLG systems).

Posts navigation

Older Posts
  • LinkedIn
  • Twitter

Top Posts & Pages

  • We need to understand what users want!
  • Bayesian vs Neural Networks
  • Blog Index
  • "Will I Pass my PhD Viva"
  • Hallucination in Neural NLG
  • Publications
  • NLG vs Templates: Levels of Sophistication in Generating Text
  • Small differences in BLEU are meaningless
  • Natural Language Generation and Machine Learning
  • Accuracy, Fluency, and Utility
Blog at WordPress.com.
Ehud Reiter's Blog
Blog at WordPress.com.
  • Follow Following
    • Ehud Reiter's Blog
    • Join 75 other followers
    • Already have a WordPress.com account? Log in now.
    • Ehud Reiter's Blog
    • Customise
    • Follow Following
    • Sign up
    • Log in
    • Report this content
    • View site in Reader
    • Manage subscriptions
    • Collapse this bar
 

Loading Comments...