In this paper, we describe the approach we designed to solve SemEval-2022 Task 8: Multilingual News Article Similarity. We collect and use exclusively textual features (title, description and body) of articles. Our best model is a stacking of 14 Transformer-based Language models fine-tuned on single or multiple fields, using data in the original language or translated to English. It placed fourth on the original leaderboard, sixth on the complete official one and fourth on the English-subset official one. We observe the data collection as our principal source of error due to a relevant fraction of missing or wrong fields.

On September 2020 a constitutional referendum was held in Italy. In this work we collect a dataset of 1.2M tweets related to this event, with particular interest to the textual content shared, and we design a hashtag-based semi-automatic approach to label them as Supporters or Against the referendum. We use the labelled dataset to train a classifier based on transformers, unsupervisedly pre-trained on Italian corpora. Our model generalizes well on tweets that cannot be labeled by the hashtag-based approach. We check that no length-, lexicon- and sentiment-biases are present to affect the performance of the classifier. Finally, we discuss the discrepancy between the magnitudes of tweets expressing a specific stance, obtained using both the hashtag-based approach and our trained classifier, and the real outcome of the referendum: the referendum was approved by 70% of the voters, while the number of tweets against the referendum is four times greater than the number of tweets supporting it. We conclude that the Italian referendum was an example of event where the minority was very loud on social media, highly influencing the perception of the event. Analyzing only the activity on social media is dangerous and can lead to extremely wrong forecasts.

Semantic sentence embeddings are usually supervisedly built minimizing distances between pairs of embeddings of sentences labelled as semantically similar by annotators. Since big labelled datasets are rare, in particular for non-English languages, and expensive, recent studies focus on unsupervised approaches that require not-paired input sentences. We instead propose a language-independent approach to build large datasets of pairs of informal texts weakly similar, without manual human effort, exploiting Twitter’s intrinsic powerful signals of relatedness: replies and quotes of tweets. We use the collected pairs to train a Transformer model with triplet-like structures, and we test the generated embeddings on Twitter NLP similarity tasks (PIT and TURL) and STSb. We also introduce four new sentence ranking evaluation benchmarks of informal texts, carefully extracted from the initial collections of tweets, proving not only that our best model learns classical Semantic Textual Similarity, but also excels on tasks where pairs of sentences are not exact paraphrases. Ablation studies reveal how increasing the corpus size influences positively the results, even at 2M samples, suggesting that bigger collections of Tweets still do not contain redundant information about semantic similarities. Code available at https://github.com/marco-digio/Twitter4SSE