Spain

Efficient document summarization requires evaluation measures that can not only rank a set of systems based on an average score, but also highlight which individual summary is better than another. However, despite the very active research on summarization approaches, few works have proposed new evaluation measures in the recent years. 

The standard measures relied upon for the development of summarization systems are most often ROUGE and BLEU which, despite being efficient in overall system ranking, remain lexical in nature and have a limited potential when it comes to training neural networks. 

In this work, we present HOLMS, a new hybrid evaluation measure for summarization that combines both language models pre-trained on large corpora and lexical similarity measures. Through several experiments, we show that HOLMS outperforms ROUGE and BLEU substantially in its correlation with human judgments on several extractive summarization datasets for both linguistic quality and pyramid scores. 


COLING 2020

HOLMS: Alternative Summary Evaluation with Large Language Models

evaluation measures

neural language models

summarization

stages 

COLING, the International Conference on Computational Linguistics, is one of the premier conferences for natural language processing and computational linguistics. Often grouped within the field of artificial intelligence, but actually pre-dating the development of artificial intelligence, advances in computational linguistics and natural language processing are now some of the major drivers behind the use of artificial intelligence for commercial and social applications – for example, on-line search, machine translation and with voice-assisted conversational devices.

First established in 1965, the biennial COLING conference is held in diverse parts of the globe and attracts participants from both top-ranked research centers and emerging countries. Today, the most important developments in our field are taking place not only in universities and academic research institutes, but also in industrial research departments and in technological startups. COLING conferences provide opportunities for all these communities to showcase their exciting developments.

COLING, the International Conference on Computational Linguistics, is one of the premier conferences for natural language processing and computational linguistics. Often grouped within the field of artificial intelligence, but actually pre-dating the development of artificial intelligence, advances in computational linguistics and natural language processing are now some of the major drivers behind the use of artificial intelligence for commercial and social applications – for example, on-line search, machine translation and with voice-assisted conversational devices.

poster

We present a new approach for detecting human-like social biases in word embeddings using representational similarity analysis. Specifically, we probe contextualized and non-contextualized embeddings for evidence of intersectional biases against Black women. We show that these embeddings represent Black women as simultaneously less feminine than White women, and less Black than Black men. This finding aligns with intersectionality theory, which argues that multiple identity categories (such as race or sex) layer on top of each other in order to create unique modes of discrimination that are not shared by any individual category.

Unequal Representations: Analyzing Intersectional Biases in Word Embeddings Using Representational Similarity Analysis

We propose a novel method that enables us to determine words that deserve to be emphasized from written text in visual media, relying only on the information from the self-attention distributions of pre-trained language models (PLMs). With extensive experiments and analyses, we show that 1) our zero-shot approach is superior to a reasonable baseline that adopts TF-IDF and that 2) there exist several attention heads in PLMs specialized for emphasis selection, confirming that PLMs are capable of recognizing important words in sentences.

Does Pre-trained Language Model Know What to Emphasize?

Maintenance record logbooks are an emerging text type in NLP. An important part of them typically consists of free text with many domain-specific technical terms, abbreviations, and non-standard spelling and grammar. This poses difficulties for NLP pipelines trained on standard corpora. Analyzing and annotating such documents is of particular importance in the development of predictive maintenance systems, which aim to improve operational efficiency, reduce costs, prevent accidents, and save lives. In order to facilitate and encourage research in this area, we have developed MaintNet, a collaborative open-source library of technical and domain-specific language resources. MaintNet provides novel logbook data from the aviation, automotive, and facility maintenance domains along with tools to aid in their (pre-)processing and clustering. Furthermore, it provides a way to encourage discussion on and sharing of new datasets and tools for logbook data analysis.

MaintNet: A Collaborative Open-Source Library for Predictive Maintenance Language Resources

This paper introduces a web demo that showcases the main characteristics of the AutoGOAL framework. AutoGOAL is a framework in Python for automatically finding the best way to solve a given task. It has been designed mainly for automatic machine learning (AutoML) but it can be used in any scenario where several possible strategies are available to solve a given computational task. In contrast with alternative frameworks, AutoGOAL can be applied seamlessly to Natural Language Processing as well as structured classification problems. This paper presents an overview of the framework’s design and experimental evaluation in several machine learning problems, including two recent NLP challenges. The accompanying software demo is available online (https://autogoal.github.io/demo) and full source code is provided under the MIT open-source license.


Demo Application for the AutoGOAL Framework

Recent advancements in NLP and machine learning have created unique challenges and opportunities for digital humanities research. In particular, there are ample opportunities for NLP and machine learning researchers to analyze data from literary texts and to broaden our understanding of human sentiment in classical Greek tragedy. In this paper, we will explore the challenges and benefits from the human and machine collaboration for sentiment analysis in Greek tragedy and address some open questions related to the collaborative annotation for the sentiments in literary texts. We focus primarily on (i) an analysis of the challenges in sentiment analysis tasks for humans and machines, and (ii) whether consistent annotation results are generated from the multiple human annotators and multiple machine annotators. For human annotators, we have used a survey-based approach with about 60 college students. We have selected three popular sentiment analysis tools for machine annotators, including VADER, CoreNLP’s sentiment annotator, and TextBlob. We have conducted a qualitative and quantitative evaluation and confirmed our observations on sentiments in Greek tragedy.

Interpretation of Sentiment Analysis in Aeschylus's Greek Tragedy

Media plays an important role in shaping public opinion. Biased media can influence people in undesirable directions and hence should be unmasked as such. We observe that feature-based and neural text classification approaches which rely only on the distribution of low-level lexical information fail to detect media bias. This weakness becomes most noticeable for articles on new events, where words appear in new contexts and hence their "bias predictiveness" is unclear. In this paper, we therefore study how second-order information about biased statements in an article helps to improve detection effectiveness. In particular, we utilize the probability distributions of the frequency, positions, and sequential order of lexical and informational sentence-level bias in a Gaussian Mixture Model. On an existing media bias dataset, we find that the frequency and positions of biased statements strongly impact article-level bias, whereas their exact sequential order is secondary. Using a standard model for sentence-level bias detection, we provide empirical evidence that article-level bias detectors that use second-order information clearly outperform those without.

Detecting Media Bias in News Articles using Gaussian Bias Distributions

We evaluate manually five lexical association measurements as regards the discovery of Modern Greek verb multiword expressions with two or more lexicalised components usingmwetoolkit3 (Ramisch et al., 2010). We use Twitter corpora and compare our findings with previous work on fiction corpora. The results of LL, MLE and T-score were found to overlap significantly in both the fiction and the Twitter corpora, while the results of PMI and Dice do not. We find that MWEs with two lexicalised components are more frequent in Twitter than in fiction corpora and that lean syntactic patterns help retrieve them more efficiently than richer ones. Our work (i) supports the enrichment of the lexicographical database for Modern Greek MWEs’IDION’ (Markantonatou et al., 2019) and (ii) highlights aspects of the usage of five association measurements on specific text genres for best MWE discovery results.

VMWE discovery: A comparative analysis between Literature and Twitter Corpora

This research investigates the collocational errors made by English learners in a learner corpus. It focuses on the extraction of unexpected collocations. A system was proposed and implemented with open source toolkit. Firstly, the collocation extraction module was evaluated by a corpus with manually annotated collocations. Secondly, a standard collocation list was collected from a corpus of native speaker. Thirdly, a list of unexpected collocations was generated by extracting candidates from a learner corpus and discarding the standard collocations on the list. The overall performance was evaluated, and possible sources of error were pointed out for future improvement.

Automatic detection of unexpected/erroneous collocations in learner corpus

The rapid advancement of technology in online communication via social media platforms has led to a prolific rise in the spread of misinformation and fake news. Fake news is especially rampant in the current COVID-19 pandemic, leading to people believing in false and potentially harmful claims and stories. Detecting fake news quickly can alleviate the spread of panic, chaos and potential health hazards. We developed a two stage automated pipeline for COVID-19 fake news detection using state of the art machine learning models for natural language processing. The first model leverages a novel fact checking algorithm that retrieves the most relevant facts concerning user claims about particular COVID-19 claims. The second model verifies the level of ``truth” in the claim by computing the textual entailment between the claim and the true facts retrieved from a manually curated COVID-19 dataset. The dataset is based on a publicly available knowledge source consisting of more than 5000 COVID-19 false claims and verified explanations, a subset of which was internally annotated and cross-validated to train and evaluate our models. We evaluate a series of models based on classical text-based features to more contextual Transformer based models and observe that a model pipeline based on BERT and ALBERT for the two stages respectively yields the best results.

Two Stage Transformer Model for COVID-19 Fake News Detection and Fact Checking

Fake news and deceptive machine-generated text are serious problems threatening modern societies, including in the Arab world. This motivates work on detecting false and manipulated stories online. However, a bottleneck for this research is lack of sufficient data to train detection models. We present a novel method for automatically generating Arabic manipulated (and potentially fake) news stories. Our method is simple and only depends on availability of true stories, which are abundant online, and a part of speech tagger (POS). To facilitate future work, we dispense with both of these requirements altogether by providing AraNews, a novel and large POS-tagged news dataset that can be used off-the-shelf. Using stories generated based on AraNews, we carry out a human annotation study that casts light on the effects of machine manipulation on text veracity. The study also measures human ability to detect Arabic machine manipulated text generated by our method. Finally, we develop the first models for detecting manipulated Arabic news and achieve state-of-the-art results on Arabic fake news detection (macro F1=70.06). Our models and data are publicly available.

HOLMS: Alternative Summary Evaluation with Large Language Models

Downloads

Next from COLING 2020

Unequal Representations: Analyzing Intersectional Biases in Word Embeddings Using Representational Similarity Analysis

Similar lecture

Multimodal Sentence Summarization via Multimodal Selective Encoding

HOLMS: Alternative Summary Evaluation with Large Language Models

.css-70qvj9{display:-webkit-box;display:-webkit-flex;display:-ms-flexbox;display:flex;-webkit-align-items:center;-webkit-box-align:center;-ms-flex-align:center;align-items:center;}Downloads

Next from COLING 2020

Unequal Representations: Analyzing Intersectional Biases in Word Embeddings Using Representational Similarity Analysis

Similar lecture

Multimodal Sentence Summarization via Multimodal Selective Encoding

Downloads