skip to main content

Title: Detecting Fine-Grained Cross-Lingual Semantic Divergences without Supervision by Learning to Rank
Detecting fine-grained differences in content conveyed in different languages matters for cross-lingual NLP and multilingual corpora analysis, but it is a challenging machine learning problem since annotation is expensive and hard to scale. This work improves the prediction and annotation of fine-grained semantic divergences. We introduce a training strategy for multilingual BERT models by learning to rank synthetic divergent examples of varying granularity. We evaluate our models on the Rationalized English-French Semantic Divergences, a new dataset released with this work, consisting of English-French sentence-pairs annotated with semantic divergence classes and token-level rationales. Learning to rank helps detect fine-grained sentence-level divergences more accurately than a strong sentence-level similarity model, while token-level predictions have the potential of further distinguishing between coarse and fine-grained divergences.
Authors:
;
Award ID(s):
1750695
Publication Date:
NSF-PAR ID:
10206137
Journal Name:
Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)
Page Range or eLocation-ID:
1563 to 1580
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract We present an event structure classification empirically derived from inferential properties annotated on sentence- and document-level Universal Decompositional Semantics (UDS) graphs. We induce this classification jointly with semantic role, entity, and event-event relation classifications using a document-level generative model structured by these graphs. To support this induction, we augment existing annotations found in the UDS1.0 dataset, which covers the entirety of the English Web Treebank, with an array of inferential properties capturing fine-grained aspects of the temporal and aspectual structure of events. The resulting dataset (available at decomp.io) is the largest annotation of event structure and (partial) event coreference to date.
  2. We present the Multilingual Amazon Reviews Corpus (MARC), a large-scale collection of Amazon reviews for multilingual text classification. The corpus contains reviews in English, Japanese, German, French, Spanish, and Chinese, which were collected between 2015 and 2019. Each record in the dataset contains the review text, the review title, the star rating, an anonymized reviewer ID, an anonymized product ID, and the coarse-grained product category (e.g., ‘books’, ‘appliances’, etc.) The corpus is balanced across the 5 possible star ratings, so each rating constitutes 20% of the reviews in each language. For each language, there are 200,000, 5,000, and 5,000 reviews in the training, development, and test sets, respectively. We report baseline results for supervised text classification and zero-shot cross-lingual transfer learning by fine-tuning a multilingual BERT model on reviews data. We propose the use of mean absolute error (MAE) instead of classification accuracy for this task, since MAE accounts for the ordinal nature of the ratings.
  3. The explosion of user-generated content (UGC)—e.g. social media posts and comments and and reviews—has motivated the development of NLP applications tailored to these types of informal texts. Prevalent among these applications have been sentiment analysis and machine translation (MT). Grounded in the observation that UGC features highly idiomatic and sentiment-charged language and we propose a decoder-side approach that incorporates automatic sentiment scoring into the MT candidate selection process. We train monolingual sentiment classifiers in English and Spanish and in addition to a multilingual sentiment model and by fine-tuning BERT and XLM-RoBERTa. Using n-best candidates generated by a baseline MT model with beam search and we select the candidate that minimizes the absolute difference between the sentiment score of the source sentence and that of the translation and and perform two human evaluations to assess the produced translations. Unlike previous work and we select this minimally divergent translation by considering the sentiment scores of the source sentence and translation on a continuous interval and rather than using e.g. binary classification and allowing for more fine-grained selection of translation candidates. The results of human evaluations show that and in comparison to the open-source MT baseline model on top of which our sentiment-basedmore »pipeline is built and our pipeline produces more accurate translations of colloquial and sentiment-heavy source texts.« less
  4. Large-scale semantic parsing datasets annotated with logical forms have enabled major advances in supervised approaches. But can richer supervision help even more? To explore the utility of fine-grained, lexical-level supervision, we introduce SQUALL, a dataset that enriches 11,276 WIKITABLEQUESTIONS English-language questions with manually created SQL equivalents plus alignments between SQL and question fragments. Our annotation enables new training possibilities for encoderdecoder models, including approaches from machine translation previously precluded by the absence of alignments. We propose and test two methods: (1) supervised attention; (2) adopting an auxiliary objective of disambiguating references in the input queries to table columns. In 5-fold cross validation, these strategies improve over strong baselines by 4.4% execution accuracy. Oracle experiments suggest that annotated alignments can support further accuracy gains of up to 23.9%.
  5. Faceted interfaces are omnipresent on the web to support data exploration and filtering. A facet is a triple: a domain (e.g., Book), a property (e.g., author, language), and a set of property values (e.g., Austen, Beauvoir, Coelho, Dostoevsky, Eco, Kerouac, Suskind, ..., French, English, German, Italian, Portuguese, Russian, ... ). Given a property (e.g., language), selecting one or more of its values (English and Italian) returns the domain entities (of type Book) that match the given values (the books that are written in English or Italian). To implement faceted interfaces in a way that is scalable to very large datasets, it is necessary to automate facet extraction. Prior work associates a facet domain with a set of homogeneous values, but does not annotate the facet property. In this paper, we annotate the facet property with a predicate from a reference Knowledge Base (KB) so as to maximize the semantic similarity between the property and the predicate. We define semantic similarity in terms of three new metrics: specificity, coverage, and frequency. Our experimental evaluation uses the DBpedia and YAGO KBs and shows that for the facet annotation problem, we obtain better results than a state-of-the-art approach for the annotation of webmore »tables as modified to annotate a set of values.« less