Sentiment analysis has been a major area of interest, for which the existence of highquality resources is crucial. In Arabic, there is a reasonable number of sentiment lexicons but with major deficiencies. The paper presents a large-scale Standard Arabic Sentiment Lexicon (SLSA) that is publicly available for free and avoids the deficiencies in the current resources. SLSA has the highest up-to-date reported coverage. The construction of SLSA is based on linking the lexicon of AraMorph with Sen-tiWordNet along with a few heuristics and powerful back-off. SLSA shows a relative improvement of 37.8% over a state-of-theart lexicon when tested for accuracy. It also outperforms it by an absolute 3.5% of F1-score when tested for sentiment analysis.

We describe a fully unsupervised cross-lingual transfer approach for part-of-speech (POS) tagging under a truly low resource scenario. We assume access to parallel translations between the target language and one or more source languages for which POS taggers are available. We use the Bible as parallel data in our experiments: small size, out-of-domain and covering many diverse languages. Our approach innovates in three ways: 1) a robust approach of selecting training instances via cross-lingual annotation projection that exploits best practices of unsupervised type and token constraints, word-alignment confidence and density of projected POS, 2) a Bi-LSTM architecture that uses contextualized word embeddings, affix embeddings and hierarchical Brown clusters, and 3) an evaluation on 12 diverse languages in terms of language family and morphological typology. In spite of the use of limited and out-of-domain parallel data, our experiments demonstrate significant improvements in accuracy over previous work. In addition, we show that using multi-source information, either via projection or output combination, improves the performance for most target languages.

The QALB-2014 shared task focuses on correcting errors in texts written in Modern Standard Arabic. In this paper, we describe the Columbia University entry in the shared task. Our system consists of several components that rely on machinelearning techniques and linguistic knowledge. We submitted three versions of the system: these share several core elements but each version also includes additional components. We describe our underlying approach and the special aspects of the different versions of our submission. Our system ranked first out of nine participating teams.

This paper describes the process of creating a novel resource, a parallel Arabizi-Arabic script corpus of SMS/Chat data. The language used in social media expresses many differences from other written genres: its vocabulary is informal with intentional deviations from standard orthography such as repeated letters for emphasis; typos and nonstandard abbreviations are common; and nonlinguistic content is written out, such as laughter, sound representations, and emoticons. This situation is exacerbated in the case of Arabic social media for two reasons. First, Arabic dialects, commonly used in social media, are quite different from Modern Standard Arabic phonologically, morphologically and lexically, and most importantly, they lack standard orthographies. Second, Arabic speakers in social media as well as discussion forums, SMS messaging and online chat often use a non-standard romanization called Arabizi. In the context of natural language processing of social media Arabic, transliterating from Arabizi of various dialects to Arabic script is a necessary step, since many of the existing state-of-the-art resources for Arabic dialect processing expect Arabic script input. The corpus described in this paper is expected to support Arabic NLP by providing this resource.

In this paper, we address the problem of converting Dialectal Arabic (DA) text that is written in the Latin script (called Arabizi) into Arabic script following the CODA convention for DA orthography. The presented system uses a finite state transducer trained at the character level to generate all possible transliterations for the input Arabizi words. We then filter the generated list using a DA morphological analyzer. After that we pick the best choice for each input word using a language model. We achieve an accuracy of 69.4% on an unseen test set compared to 63.1% using a system which represents a previously proposed approach.