A Massively Multilingual Analysis of Cross-linguality in Shared Embedding Space

In cross-lingual language models, representations for many different languages live in the same space. Here, we investigate the linguistic and non-linguistic factors affecting sentence-level alignment in cross-lingual pretrained language models for 101 languages and 5,050 language pairs. Using BERT-based LaBSE and BiLSTM-based LASER as our models, and the Bible as our corpus, we compute a task-based measure of cross-lingual alignment in the form of bitext retrieval performance, as well as four intrinsic measures of vector space alignment and isomorphism. We then examine a range of linguistic, quasi-linguistic, and training-related features as potential predictors of these alignment metrics. The results of our analyses show that word order agreement and agreement in morphological complexity are two of the strongest linguistic predictors of cross-linguality. We also note in-family training data as a stronger predictor than language-specific training data across the board. We verify some of our linguistic findings by looking at the effect of morphological segmentation on English-Inuktitut alignment, in addition to examining the effect of word order agreement on isomorphism for 66 zero-shot language pairs from a different corpus. We make the data and code for our experiments publicly available.

How Does Fine-tuning Affect the Geometry of Embedding Space: A Case Study on Isotropy

It is widely accepted that fine-tuning pre-trained language models usually brings about performance improvements in downstream tasks. However, there are limited studies on the reasons behind this effectiveness, particularly from the viewpoint of structural changes in the embedding space. Trying to fill this gap, in this paper, we analyze the extent to which the isotropy of the embedding space changes after fine-tuning. We demonstrate that, even though isotropy is a desirable geometrical property, fine-tuning does not necessarily result in isotropy enhancements. Moreover, local structures in pre-trained contextual word representations (CWRs), such as those encoding token types or frequency, undergo a massive change during fine-tuning. Our experiments show dramatic growth in the number of elongated directions in the embedding space, which, in contrast to pre-trained CWRs, carry the essential linguistic knowledge in the fine-tuned embedding space, making existing isotropy enhancement methods ineffective.

Char2Subword: Extending the Subword Embedding Space Using Robust Character Compositionality

Byte-pair encoding (BPE) is a ubiquitous algorithm in the subword tokenization process of language models as it provides multiple benefits. However, this process is solely based on pre-training data statistics, making it hard for the tokenizer to handle infrequent spellings. On the other hand, though robust to misspellings, pure character-level models often lead to unreasonably long sequences and make it harder for the model to learn meaningful words. To alleviate these challenges, we propose a character-based subword module (char2subword) that learns the subword embedding table in pre-trained models like BERT. Our char2subword module builds representations from characters out of the subword vocabulary, and it can be used as a drop-in replacement of the subword embedding table. The module is robust to character-level alterations such as misspellings, word inflection, casing, and punctuation. We integrate it further with BERT through pre-training while keeping BERT transformer parameters fixed--and thus, providing a practical method. Finally, we show that incorporating our module to mBERT significantly improves the performance on the social media linguistic code-switching evaluation (LinCE) benchmark.

APo-VAE: Text Generation in Hyperbolic Space

Natural language often exhibits inherent hierarchical structure ingrained with complex syntax and semantics. However, most state-of-the-art deep generative models learn embeddings only in Euclidean vector space, without accounting for this structural property of language. In this paper, we investigate text generation in a hyperbolic latent space to learn continuous hierarchical representations. An Adversarial Poincare Variational Autoencoder (APo-VAE) is presented, where both the prior and variational posterior of latent variables are defined over a Poincare ball via wrapped normal distributions. By adopting the primal-dual formulation of Kullback-Leibler divergence, an adversarial learning procedure is introduced to empower robust model training. Extensive experiments in language modeling, unaligned style transfer, and dialog-response generation demonstrate the effectiveness of the proposed APo-VAE model over VAEs in Euclidean latent space, thanks to its superb capabilities in capturing latent language hierarchies in hyperbolic space.

Smoothing and Shrinking the Sparse Seq2Seq Search Space

Current sequence-to-sequence models are trained to minimize cross-entropy and use softmax to compute the locally normalized probabilities over target sequences. While this setup has led to strong results in a variety of tasks, one unsatisfying aspect is its length bias: models give high scores to short, inadequate hypotheses and often make the empty string the argmax---the so-called cat got your tongue problem. Recently proposed entmax-based sparse sequence-to-sequence models present a possible solution, since they can shrink the search space by assigning zero probability to bad hypotheses, but their ability to handle word-level tasks with transformers has never been tested. In this work, we show that entmax-based models effectively solve the cat got your tongue problem, removing a major source of model error for neural machine translation. In addition, we generalize label smoothing, a critical regularization technique, to the broader family of Fenchel-Young losses, which includes both cross-entropy and the entmax losses. Our resulting label-smoothed entmax loss models set a new state of the art on multilingual grapheme-to-phoneme conversion and deliver improvements and better calibration properties on cross-lingual morphological inflection and machine translation for 7 language pairs.

Learning to Lemmatize in the Word Representation Space

Lemmatization is often used with morphologically rich languages to address issues caused by morphological complexity, performed by grammar-based lemmatizers. We propose an alternative for this, in form of a tool that performs lemmatization in the space of word embeddings. Word embeddings as distributed representations natively encode some information about the relationship between base and inflected forms, and we show that it is possible to learn a transformation that approximately maps the embeddings of inflected forms to the embeddings of the corresponding lemmas. This facilitates an alternative processing pipeline that replaces traditional lemmatization with the lemmatizing transformation in downstream processing for any application. We demonstrate the method in the Finnish language, outperforming traditional lemmatizers in example task of document similarity comparison, but the approach is language independent and can be trained for new languages with mild requirements.

Understanding the Semantic Space: How Word Meanings Dynamically Adapt in the Context of a Sentence

How do people understand the meaning of the word small'' when used to describe a mosquito, a church, or a planet? While humans have a remarkable ability to form meanings by combining existing concepts, modeling this process is challenging. This paper addresses that challenge through CEREBRA (Context-dEpendent meaning REpresentations in the BRAin) neural network model. CEREBRA characterizes how word meanings dynamically adapt in the context of a sentence by decomposing sentence fMRI into words and words into embodied brain-based semantic features. It demonstrates that words in different contexts have different representations and the word meaning changes in a way that is meaningful to human subjects. CEREBRA's context-based representations can potentially be used to make NLP applications more human-like.

Parameter Space Factorization for Zero-Shot Learning across Tasks and Languages

Abstract Most combinations of NLP tasks and language varieties lack in-domain examples for supervised training because of the paucity of annotated data. How can neural models make sample-efficient generalizations from task--language combinations with available data to low-resource ones? In this work, we propose a Bayesian generative model for the space of neural parameters. We assume that this space can be factorized into latent variables for each language and each task. We infer the posteriors over such latent variables based on data from seen task--language combinations through variational inference. This enables zero-shot classification on unseen combinations at prediction time. For instance, given training data for named entity recognition (NER) in Vietnamese and for part-of-speech (POS) tagging in Wolof, our model can perform accurate predictions for NER in Wolof. In particular, we experiment with a typologically diverse sample of 33 languages from 4 continents and 11 families, and show that our model yields comparable or better results than state-of-the-art, zero-shot cross-lingual transfer methods. Our code is available at github.com/cambridgeltl/parameter-factorization.

Analytical study of urban networks at the level of spatial development “Greater Damascus Region as case study”

With increased attention to regional and spatial planning in the world and the need to pay attention to modern spatial development methods, increased thinking for necessity urban networks as a mechanism for achieving spatial development in small and medium-sized cities, operating alone, has been enhanced by changing the spatial structure of the Territory, raising qualitative and quantitative use, To collect the assets and resources of the cities of the Region that accept participation in potential opportunities and risks. The analytical study was implemented by proposing an urban network in the Greater Damascus Region that examines the nodes and links, and uses the SWOT system to examine opportunities and possibilities, its relationship to the problems and determinants of selecting the best scenario for the proposed urban network examines the effect of applying the urban network concept on changing the spatial structure of the Greater Damascus Region, and propose a workable mechanism at the spatial-sectorial level. The research concludes with a set of conclusions and recommendations that determine the importance of the networking of cities according to their location, which is capable of adapting to the changes taking place in our Syrian cities. In the absence of a future vision for regional development that is appropriate to the current situation, taking into account the economic, social and spatial changes taking place; and which hinder the achievement of the proposed national framework for regional planning in 2010.

The Relationship between social adjustment and scholastic achievement (A field study of a sample of children in the childfriendly space In Jermana)

The current research aims to identify the relationship between social adjustment of children and their degree of achievement. The impact of the sex variable is known in a segment of children whose families were displaced during the period of the Syrian crisis from the Damascus countryside to the city of Jaramana. The descriptive analytical approach was used in the field study and sample survey, which consisted of 011child-friendly children and children in the city of Jaramana. The research tool consisted of a questionnaire to measure the social compatibility and educational achievement of children., The research has reached a number of results, the most important of which are: -There is no significant correlation between the social adjustment of the sample and the academic achievement. -There was a statistically significant correlation between the social adjustment of female sample members and the academic achievement. -There was no statistically significant correlation between social adjustment of male sample members and academic achievement.