Did AI get more negative recently?

Abstract

In this paper, we classify scientific articles in the domain of natural language processing (NLP) and machine learning (ML), as core subfields of artificial intelligence (AI), into whether (i) they extend the current state-of-the-art by the introduction of novel techniques which beat existing models or whether (ii) they mainly criticize the existing state-of-the-art, i.e. that it is deficient with respect to some property (e.g. wrong evaluation, wrong datasets, misleading task specification). We refer to contributions under (i) as having a ‘positive stance’ and contributions under (ii) as having a ‘negative stance’ (to related work). We annotate over 1.5 k papers from NLP and ML to train a SciBERT-based model to automatically predict the stance of a paper based on its title and abstract. We then analyse large-scale trends on over 41 k papers from the last approximately 35 years in NLP and ML, finding that papers have become substantially more positive over time, but negative papers also got more negative and we observe considerably more negative papers in recent years. Negative papers are also more influential in terms of citations they receive.

Transformers Go for the LOLs: Generating (Humourous) Titles from Scientific Abstracts End-to-End

“…We use the dataset released by Beese et al (2023), which contains title-abstract pairs and corresponding meta-information such as the publication year and venue. Beese et al (2023) extracted the data from two sources: ACL Anthology (from 1984 to 2021) and machine learning conferences (from 1989 to 2021); we refer to the datasets from these two sources as NLP and ML, respectively. After filtering (described in Appendix A), 32,952 abstracttitle pairs remain in our dataset.…”

“…Stage 2: To find more funny title candidates to annotate, the two annotators annotated the funniest 396 titles in the original dataset from Beese et al (2023), predicted by the Stage 1 ensemble classifier; 75.8% (300 titles) were judged as FUNNY or FUNNY med , which is substantially higher than the proportion of funny titles in the annotated data of Stage 1 (7.3%). Thus, the annotated data expands to 2,441 titles (= 1, 730 + 315 + 396), where 1,893 are labeled as ¬FUNNY, 492 as FUNNY med and 56 as FUNNY.…”

Utilization of Transformer-Based Language Models in Understanding Citizens’ Interests, Sentiments and Emotions Towards Public Services Digitalization

“…Such models could be applied in completely unsupervised settings. One novel pre-trained model for sentiment analysis is SiEBERT which potential is demonstrated in [16,17,18,19].…”