Dialectal Arabic to Modern Standard Arabic (DA-MSA) translation is a challenging task in Machine Translation (MT) due to significant lexical, syntactic, and semantic divergences between Arabic dialects and MSA. Existing automatic evaluation metrics and general-purpose human evaluation frameworks struggle to capture dialect-specific MT errors, hindering progress in translation assessment. This paper introduces Ara-HOPE, a human-centric post-editing evaluation framework designed to systematically address these challenges. The framework includes a five-category error taxonomy and a decision-tree annotation protocol. Through comparative evaluation of three MT systems (Arabic-centric Jais, general-purpose GPT-3.5, and baseline NLLB-200), Ara-HOPE effectively highlights systematic performance differences between these systems. Our results show that dialect-specific terminology and semantic preservation remain the most persistent challenges in DA-MSA translation. Ara-HOPE establishes a new framework for evaluating Dialectal Arabic MT quality and provides actionable guidance for improving dialect-aware MT systems. For reproducibility, we make the annotation files and related materials publicly available at https://github.com/abdullahalabdullah/Ara-HOPE.

A large part of political campaigns during elections is now being conducted online, with political actors leveraging their networks on social media platforms. To maintain transparency in political communications, regulations applicable to online campaigning have been put in place in many democracies. While it should be straightforward for voters to determine who produced and funded online advertisements comprising paid political campaigns, it is much more challenging to detect if organic content, i.e., social media posts, pertains to political campaigning, due to possibly subtle yet suggestive language that can be used by certain actors. In this paper, we investigate the feasibility of automatically detecting whether a given tweet posted by a political actor pertains to political campaigning, and if yes, whether it was conveyed in a direct or indirect (subtle) manner. After establishing an annotation scheme for the task of detecting political campaign content in tweets, we fine-tuned three encoder models (BERT, BERTweet and PoliBERTweet) for the same task and evaluated their performance. Our results show that fine-tuning BERTweet leads to the best macro-averaged F1-score (0.776), although all models consistently struggle to detect indirect campaigning.