Papers › Analysis of the Evolution of Advanced Transformer-Based Language Models: Experiments...

Analysis of the Evolution of Advanced Transformer-Based Language Models: Experiments on Opinion Mining

7 Aug 2023arXiv:2308.03235archive 2025-07-28

Nour Eddine Zekaoui, Siham Yousfi, Maryem Rhanoui, Mounia Mikram

Opinion mining, also known as sentiment analysis, is a subfield of natural language processing (NLP) that focuses on identifying and extracting subjective information in textual material. This can include determining the overall sentiment of a piece of text (e.g., positive or negative), as well as identifying specific emotions or opinions expressed in the text, that involves the use of advanced machine and deep learning techniques. Recently, transformer-based language models make this task of human emotion analysis intuitive, thanks to the attention mechanism and parallel computation. These advantages make such models very powerful on linguistic tasks, unlike recurrent neural networks that spend a lot of time on sequential processing, making them prone to fail when it comes to processing long text. The scope of our paper aims to study the behaviour of the cutting-edge Transformer-based language models on opinion mining and provide a high-level comparison between them to highlight their key particularities. Additionally, our comparative study shows leads and paves the way for production engineers regarding the approach to focus on and is useful for researchers as it provides guidelines for future research subjects.

PaperPDFCode

Code

zekaouinoureddine/Opinion-Transformers officialmentioned in paperpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Emotion RecognitionOpinion MiningSentiment AnalysisSentiment ClassificationText Classification

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Opinion Mining IMDb Movie Reviews ELECTRA Accuracy 95.6 #1 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews ELECTRA F1 95.6 #1 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews RoBERTa Accuracy 95.3 #2 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews RoBERTa F1 95.3 #2 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews DeBERTa Accuracy 95.1 #3 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews DeBERTa F1 95.1 #3 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews Longformer Accuracy 95.0 #4 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews XLNet Accuracy 94.8 #5 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews XLNet F1 94.9 #5 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews BART Accuracy 94.6 #6 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews ConvBERT Accuracy 94.5 #7 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews ConvBERT F1 94.6 #7 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews BERT Accuracy 94.0 #8 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews BERT F1 94.1 #8 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews T5 Accuracy 93.9 #9 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews T5 F1 94.0 #9 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews DistilBERT Accuracy 93.4 #10 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews DistilBERT F1 93.5 #10 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews ALBERT Accuracy 93.0 #11 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews ALBERT F1 93.0 #11 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews XLM-RoBERTA Accuracy 75.2 #12 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews XLM-RoBERTA F1 77.0 #12 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews GPT-2 Accuracy 54.5 #13 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews GPT-2 F1 52.9 #13 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews GPT Accuracy 53.2 #14 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews GPT F1 66.4 #14 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews Reformer Accuracy 52.2 #15 of 15 Archive leaderboard report
Opinion Mining IMDb Movie Reviews Reformer F1 53.3 #15 of 15 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Methods

AdamBERTDropoutELECTRALayer NormalizationLinear LayerLinear Warmup With Linear DecayMulti-Head AttentionResidual ConnectionRoBERTaSoftmaxWeight DecayWordPiecefail

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections