Better Summarization Evaluation with Word Embeddings for ROUGE

08/25/2015
by   Jun-Ping Ng, et al.
0

ROUGE is a widely adopted, automatic evaluation measure for text summarization. While it has been shown to correlate well with human judgements, it is biased towards surface lexical similarities. This makes it unsuitable for the evaluation of abstractive summarization, or summaries with substantial paraphrasing. We study the effectiveness of word embeddings to overcome this disadvantage of ROUGE. Specifically, instead of measuring lexical overlaps, word embeddings are used to compute the semantic similarity of the words used in summaries instead. Our experimental results show that our proposal is able to achieve better correlations with human judgements when measured with the Spearman and Kendall rank coefficients.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
10/20/2017

A Semantically Motivated Approach to Compute ROUGE Scores

ROUGE is one of the first and most widely used evaluation metrics for te...
research
04/01/2016

Revisiting Summarization Evaluation for Scientific Articles

Evaluation of text summarization approaches have been mostly based on me...
research
08/27/2019

Facet-Aware Evaluation for Extractive Text Summarization

Commonly adopted metrics for extractive text summarization like ROUGE fo...
research
01/21/2018

A Survey of Word Embeddings Evaluation Methods

Word embeddings are real-valued word representations able to capture lex...
research
04/14/2020

Extending Text Informativeness Measures to Passage Interestingness Evaluation (Language Model vs. Word Embedding)

Standard informativeness measures used to evaluate Automatic Text Summar...
research
05/25/2020

AutoSUM: Automating Feature Extraction and Multi-user Preference Simulation for Entity Summarization

Withthegrowthofknowledgegraphs, entity descriptions are becoming extreme...
research
07/31/2020

Evaluating Semantic Interaction on Word Embeddings via Simulation

Semantic interaction (SI) attempts to learn the user's cognitive intents...

Please sign up or login with your details

Forgot password? Click here to reset