1,720,992 research outputs found
Hiding Your Face Is Not Enough: user identity linkage with image recognition
No full textPeople tend to have multiple identities or personalities in theirreal and on-linelives. In the real life, these identities can be even associated with different names used with parents, groups of friends or in formal contexts. In the on-line side of life, the attitude has exploded: people have the possibility to express different identities with different names in different social networks (SNs), interfacing with these tools claiming the same meaning as the actions and connections in real life. Thus, a fundamental question arises-Can profiles of the same user be connected in multiple SNs?In this paper, we present Hiding Your Face Is Not Enough (HYFINE) model: a User Identity Linking model that fully exploits images in profiles. Our HYFINE model consists of two parts: (1) the corpus extraction system; (2) the classification systemHYFINE-c, which classify if two profiles to determine if these profiles are two different identities of the same user by fully using images along with other features. We show that HYFINE model, exploiting images in profiles, can match profiles of the users in different SNs with high performance
Dis-cover ai minds to preserve human knowledge
No full textModern AI technologies make use of statistical learners that lead to self-empiricist logic, which, unlike human minds, use learned non-symbolic representations. Nevertheless, it seems that it is not the right way to progress in AI. The structure of symbols—the operations by which the intellectual solution is realized—and the search for strategic reference points evoke important issues in the analysis of AI. Studying how knowledge can be represented through methods of theoretical generalization and empirical observation is only the latest step in a long process of evolution. For many years, humans, seeing language as innate, have carried out symbolic theories. Everything seems to have skipped ahead with the advent of Machine Learning. In this paper, after a long analysis of history, the rule-based and the learning-based vision, we would investigate the syntax as possible meeting point between the different learning theories. Finally, we propose a new vision of knowledge in AI models based on a combination of rules, learning, and human knowledge
KERMITviz: Visualizing Neural Network Activations on Syntactic Trees
No full textThe study of symbolic syntactic interpretations has been the cornerstone of natural language understanding for many years. Today, modern artificial neural networks are widely searched to assess their syntactic ability, through several probing tasks. In this paper, we propose a neural network system that explicitly includes syntactic interpretations: Kernel-inspired Encoder with Recursive Mechanism for Interpretable Trees Visualizer (KERMITviz). The most important result is that KERMITviz allows to visualize how syntax is used in inference. This system can be used in combination with transformer architectures like BERT, XLNet and clarifies the use of symbolic syntactic interpretations in specific neural networks making the black-box neural network neural networks explainable, interpretable and clear
Syntax and prejudice: ethically-charged biases of a syntax-based hate speech recognizer unveiled
No full textHate speech recognizers (HSRs) can be the panacea for containing hate in social media or can result in the biggest form of prejudice-based censorship hindering people to express their true selves. In this paper, we hypothesized how massive use of syntax can reduce the prejudice effect in HSRs. To explore this hypothesis, we propose Unintended-bias Visualizer based on Kermit modeling (KERM-HATE): a syntax-based HSR, which is endowed with syntax heat parse trees used as a post-hoc explanation of classifications. KERM-HATE significantly outperforms BERT-based, RoBERTa-based and XLNet-based HSR on standard datasets. Surprisingly this result is not sufficient. In fact, the post-hoc analysis on novel datasets on recent divisive topics shows that even KERM-HATE carries the prejudice distilled from the initial corpus. Therefore, although tests on standard datasets may show higher performance, syntax alone cannot drive the ‘‘attention“ of HSRs to ethically-unbiased features
Shedding Light on the Dark Web: Authorship Attribution in Radical Forums
No full textOnline users tend to hide their real identities by adopting different names on the Internet. On Facebook or LinkedIn, for example, people usually appear with their real names. On other standard websites, such as forums, people often use nicknames to protect their real identities. Aliases are used when users are trying to protect their anonymity. This can be a challenge to law enforcement trying to identify users who often change nicknames. In unmonitored contexts, such as the dark web, users expect strong identity protection. Thus, without censorship, these users may create parallel social networks where they can engage in potentially malicious activities that could pose security threats. In this paper, we propose a solution to the need to recognize people who anonymize themselves behind nicknames—the authorship attribution (AA) task—in the challenging context of the dark web: specifically, an English-language Islamic forum dedicated to discussions of issues related to the Islamic world and Islam, in which members of radical Islamic groups are present. We provide extensive analysis by testing models based on transformers, styles, and syntactic features. Downstream of the experiments, we show how models that analyze syntax and style perform better than pre-trained universal language models
Empowering cross-lingual abilities of instruction-tuned large language models by translation-following demonstrations
Full text linkThe language ability of Large Language Models (LLMs) is often unbalanced towards English because of the imbalance in the distribution of the pre-training data. This disparity is demanded in further fine-tuning and affecting the cross-lingual abilities of LLMs. In this paper, we propose to empower Instruction-tuned LLMs (It-LLMs) in languages other than English by building semantic alignment between them. Hence, we propose CrossAl-paca, an It-LLM with cross-lingual Instruction-following and Translation-following demon-strations to improve semantic alignment between languages. We validate our approach on the multilingual Question Answering (QA) benchmarks XQUAD and MLQA and adapted versions of MMLU and BBH. Our models, tested over six different languages, outperform the It-LLMs tuned on monolingual data. The final results show that instruction tuning on non-English data is not enough and that semantic alignment can be further improved by Translation-following demonstrations.<br/
- …
