Multimodal learning pipelines have benefited from the success of pretrained language models. However, this comes at the cost of increased model parameters. In this work, we propose Adapted Multimodal BERT (AMB), a BERT-based architecture for multimodal tasks that uses a combination of adapter modules and intermediate fusion layers. The adapter adjusts the pretrained language model for the task at hand, while the fusion layers perform task-specific, layer-wise fusion of audio-visual information with textual BERT representations. During the adaptation process the pre-trained language model parameters remain frozen, allowing for fast, parameter-efficient training. In our ablations we see that this approach leads to efficient models, that can outperform their fine-tuned counterparts and are robust to input noise. Our experiments on sentiment analysis with CMU-MOSEI show that AMB outperforms the current state-of-the-art across metrics, with 3.4% relative reduction in the resulting error and 2.1% relative improvement in 7-class classification accuracy.

Open Information Extraction (OpenIE) facilitates the open-domain discovery of textual facts. However, the prevailing solutions evaluate OpenIE models on in-domain test sets aside from the training corpus, which certainly violates the initial task principle of domain-independence. In this paper, we propose to advance OpenIE towards a more realistic scenario: generalizing over unseen target domains with different data distributions from the source training domains, termed Generalized OpenIE. For this purpose, we first introduce GLOBE, a large-scale human-annotated multi-domain OpenIE benchmark, to examine the robustness of recent OpenIE models to domain shifts, and the relative performance degradation of up to 70% implies the challenges of generalized OpenIE. Then, we propose DragonIE, which explores a minimalist graph expression of textual fact: directed acyclic graph, to improve the OpenIE generalization. Extensive experiments demonstrate that DragonIE beats the previous methods in both in-domain and out-of-domain settings by as much as 6.0% in F1 score absolutely, but there is still ample room for improvement.

Meta has been hit with a €265 million ($277 million) fine for failing to prevent millions of Facebook users' mobile phone numbers and other data from being scraped and dumped online, Independent.ie It's the second fine levied by the Irish Data Protection Commission (DPC) in just the past few months, following a €405 million ($402 million at the time) penalty issued in September. In just the last 18 months, Meta has tallied nearly €1 billion in fines. The penalty was issued in response to the leak of 533 million Facebook users' data reported in April last year. That included phone numbers, birth dates, email addresses and locations, information that could be exploited in phishing and other attacks.

Earlier this year, Twitter confirmed that the private user data for 5.4 million users was stolen due to an API vulnerability, but the company said it had "no evidence" that it was exploited. Now, all of those accounts have been exposed on a hacker form, BleepingComputer has reported. On top of that, an additional 1.4 million Twitter profiles for suspended users was reportedly shared privately, and an even larger data dump with the data of "tens of millions" of other users may have come from the same vulnerability. The owner of hacking forum called Breached told BleepingComputer that it was responsible for exploiting the weakness (originally obtained from another hacker called "Devil") and dumping the user records. It said that it also obtained 1.4 million Twitter profiles for suspended accounts, obtained via another API, but only shared those privately among a few individuals.

In recent years, multilingual pre-trained language models have gained prominence due to their remarkable performance on numerous downstream Natural Language Processing tasks (NLP). However, pre-training these large multilingual language models requires a lot of training data, which is not available for African Languages. Active learning is a semi-supervised learning algorithm, in which a model consistently and dynamically learns to identify the most beneficial samples to train itself on, in order to achieve better optimization and performance on downstream tasks. Furthermore, active learning effectively and practically addresses real-world data scarcity. Despite all its benefits, active learning, in the context of NLP and especially multilingual language models pretraining, has received little consideration. In this paper, we present AfroLM, a multilingual language model pretrained from scratch on 23 African languages (the largest effort to date) using our novel self-active learning framework. Pretrained on a dataset significantly (14x) smaller than existing baselines, AfroLM outperforms many multilingual pretrained language models (AfriBERTa, XLMR-base, mBERT) on various NLP downstream tasks (NER, text classification, and sentiment analysis). Additional out-of-domain sentiment analysis experiments show that \textbf{AfroLM} is able to generalize well across various domains. We release the code source, and our datasets used in our framework at https://github.com/bonaventuredossou/MLM_AL.

Aspect-based sentiment analysis (ABSA) predicts sentiment polarity towards a specific aspect in the given sentence. While pre-trained language models such as BERT have achieved great success, incorporating dynamic semantic changes into ABSA remains challenging. To this end, in this paper, we propose to address this problem by Dynamic Re-weighting BERT (DR-BERT), a novel method designed to learn dynamic aspect-oriented semantics for ABSA. Specifically, we first take the Stack-BERT layers as a primary encoder to grasp the overall semantic of the sentence and then fine-tune it by incorporating a lightweight Dynamic Re-weighting Adapter (DRA). Note that the DRA can pay close attention to a small region of the sentences at each step and re-weigh the vitally important words for better aspect-aware sentiment understanding. Finally, experimental results on three benchmark datasets demonstrate the effectiveness and the rationality of our proposed model and provide good interpretable insights for future semantic modeling.

The president of the European Commission, the executive branch of the European Union, has confirmed there are multiple ongoing investigations into TikTok. The probes concern the transfer of EU citizens' data to China and targeted advertising aimed at minors. Investigators are seeking to ensure that TikTok meets General Data Protection Regulation ( GDPR) requirements. "The data practices of TikTok, including with respect to international data transfers, are the object of several ongoing proceedings," Ursula von der Leyden wrote in a letter shared by Federal Communications Commissioner Brendan Carr. "This includes an investigation by the Irish [Data Protection Commission] about TikTok's compliance with several GDPR requirements, including as regards data transfers to China and the processing of data of minors, and litigation before the Dutch courts (in particular concerning targeted advertising regarding minors and data transfers to China)."

The automatic extraction of materials and related properties from the scientific literature is gaining attention in data-driven materials science (Materials Informatics). In this paper, we discuss Grobid-superconductors, our solution for automatically extracting superconductor material names and respective properties from text. Built as a Grobid module, it combines machine learning and heuristic approaches in a multi-step architecture that supports input data as raw text or PDF documents. Using Grobid-superconductors, we built SuperCon2, a database of 40324 materials and properties records from 37700 papers. The material (or sample) information is represented by name, chemical formula, and material class, and is characterized by shape, doping, substitution variables for components, and substrate as adjoined information. The properties include the Tc superconducting critical temperature and, when available, applied pressure with the Tc measurement method.

Detecting opportunities and threats from massive text data is a challenging task for most. Traditionally, companies would rely mainly on structured data to detect and predict risks, losing a huge amount of information that could be extracted from unstructured text data. Fortunately, artificial intelligence came to remedy this issue by innovating in data extraction and processing techniques, allowing us to understand and make use of Natural Language data and turning it into structures that a machine can process and extract insight from. Uncertainty refers to a state of not knowing what will happen in the future. This paper aims to leverage natural language processing and machine learning techniques to model uncertainties and evaluate the risk level in each uncertainty cluster using massive text data.

Washington – Twitter's owner Elon Musk has pledged the platform will not become a "hellscape," but experts fear a staff exodus following mass layoffs may have devastated its ability to combat misinformation, impersonation and data theft. Twitter devolved into what campaigners described as a cesspit of falsehoods and hate speech after recent layoffs cut half the company's 7,500 staff and fake accounts proliferated following its botched rollout of a paid verification system. This could be due to a conflict with your ad-blocking or security software. Please add japantimes.co.jp and piano.io to your list of allowed sites. If this does not resolve the issue or you are unable to add the domains to your allowlist, please see this FAQ.