Żeljko Agić

Inverted indexing for cross-lingual NLP

We present a novel, count-based approach to obtaining inter-lingual word representations based on inverted indexing of Wikipedia. We present experiments applying these representations to 17 datasets in document classification, POS tagging, dependency parsing, and word alignment. Our approach has the advantage that it is simple, computationally efficient and almost parameter-free, and, more importantly, it enables multi-source crosslingual learning. In 14/17 cases, we improve over using state-of-the-art bilingual embeddings.

If all you have is a bit of the Bible: Learning POS taggers for truly low-resource languages

We present a simple method for learning part-of-speech taggers for languages like Akawaio, Aukan, or Cakchiquel – languages for which nothing but a translation of parts of the Bible exists. By aggregating over the tags from a few annotated languages and spreading them via wordalignment on the verses, we learn POS taggers for 100 languages, using the languages to bootstrap each other. We evaluate our cross-lingual models on the 25 languages where test sets exist, as well as on another 10 for which we have tag dictionaries. Our approach performs much better (20-30%) than state-of-the-art unsupervised POS taggers induced from Bible translations, and is often competitive with weakly supervised approaches that assume high-quality parallel corpora, representative monolingual corpora with perfect tokenization, and/or tag dictionaries. We make models for all 100 languages available.

Treebank Translation for Cross-Lingual Parser Induction

Cross-lingual learning has become a popular approach to facilitate the development of resources and tools for low density languages. Its underlying idea is to make use of existing tools and annotations in resource-rich languages to create similar tools and resources for resource-poor languages. Typically, this is achieved by either projecting annotations across parallel corpora, or by transferring models from one or more source languages to a target language. In this paper, we explore a third strategy by using machine translation to create synthetic training data from the original source-side annotations. Specifically, we apply this technique to dependency parsing, using a cross-lingually unified treebank for adequate evaluation. Our approach draws on annotation projection but avoids the use of noisy source-side annotation of an unrelated parallel corpus and instead relies on manual treebank annotation in combination with statistical machine translation, which makes it possible to train fully lexicalized parsers. We show that this approach significantly outperforms delexicalized transfer parsing.% despite the error-prone translation step.

Potsdam: Semantic Dependency Parsing by Bidirectional Graph-Tree Transformations and Syntactic Parsing

We present the Potsdam systems that participated in the semantic dependency parsing shared task of SemEval 2014. They are based on linguistically motivated bidirectional transformations between graphs and trees and on utilization of syntactic dependency parsing. They were entered in both the closed track and the open track of the challenge, recording a peak average labeled F1 score of 78.60.

Cross-lingual Dependency Parsing of Related Languages with Rich Morphosyntactic Tagsets

This paper addresses cross-lingual dependency parsing using rich morphosyntactic tagsets. In our case study, we experiment with three related Slavic languages: Croatian, Serbian and Slovene. Four different dependency treebanks are used for monolingual parsing, direct cross-lingual parsing, and a recently introduced crosslingual parsing approach that utilizes statistical machine translation and annotation projection. We argue for the benefits of using rich morphosyntactic tagsets in cross-lingual parsing and empirically support the claim by showing large improvements over an impoverished common feature representation in form of a reduced part-of-speech tagset. In the process, we improve over the previous state-of-the-art scores in dependency parsing for all three languages.

Do dependency parsing metrics correlate with human judgments

Using automatic measures such as labeled and unlabeled attachment scores is common practice in dependency parser evaluation. In this paper, we examine whether these measures correlate with human judgments of overall parse quality. We ask linguists with experience in dependency annotation to judge system outputs. We measure the correlation between their judgments and a range of parse evaluation metrics across five languages. The humanmetric correlation is lower for dependency parsing than for other NLP tasks. Also, inter-annotator agreement is sometimes higher than the agreement between judgments and metrics, indicating that the standard metrics fail to capture certain aspects of parse quality, such as the relevance of root attachment or the relative importance of the different parts of speech.