While important properties of word vector representations have been studied extensively, far less is known about the properties of sentence vector representations. Word vectors are often evaluated by assessing to what degree they exhibit regularities with regard to relationships of the sort considered in word analogies. In this paper, we investigate to what extent commonly used sentence vector representation spaces as well reflect certain kinds of regularities. We propose a number of schemes to induce evaluation data, based on lexical analogy data as well as semantic relationships between sentences. Our experiments consider a wide range of sentence embedding methods, including ones based on BERT-style contextual embeddings. We find that different models differ substantially in their ability to reflect such regularities.
Cognitive Machine Learning (ENS - CNRS - EHESS - INRIA - PSL Research University)（认知机器学习（巴黎高等师范学院-法国国家科学研究中心-法国社会科学高等研究院-法国国家信息与自动化研究所-巴黎文理研究大学））；Facebook A.I. Research（Facebook人工智能研究院）；Laboratoire de Linguistique Formelle (CNRS - Paris Diderot - Sorbonne Paris Cite)（形式语言学实验室（法国国家科学研究中心-巴黎狄德罗大学-巴黎索邦大学））；University of Toronto（多伦多大学）
Vector space models of words have long been claimed to capture linguistic regularities as simple vector translations, but problems have been raised with this claim. We decompose and empirically analyze the classic arithmetic word analogy test, to motivate two new metrics that address the issues with the standard test, and which distinguish between class-wise offset concentration (similar directions between pairs of words drawn from different broad classes, such as France--London, China--Ottawa, ...) and pairing consistency (the existence of a regular transformation between correctly-matched pa…
University of Massachusetts Amherst（马萨诸塞大学阿默斯特分校）；IBM Research（IBM研究院）
Recent work has explored methods for learning continuous vector space word representations reflecting the underlying semantics of words. Simple vector space arithmetic using cosine distances has been shown to capture certain types of analogies, such as reasoning about plurals from singulars, past tense from present tense, etc. In this paper, we introduce a new approach to capture analogies in continuous word representations, based on modeling not just individual word vectors, but rather the subspaces spanned by groups of words. We exploit the property that the set of subspaces in n-dimensional…
We experiment with new methods for learning how related words are positioned relative to each other in word embedding spaces. Previous approaches learned constant vector offsets: vectors that point from source tokens to target tokens with an assumption that these offsets were parallel to each other. We show that the offsets between related tokens are closer to orthogonal than parallel, and that they have low cosine similarities. We proceed by making a different assumption; target tokens are linearly separable from source and un-labeled tokens. We show that a max-margin hyperplane can separate …
CNRS-EHESS-INRIA（法国国家科学研究中心-法国社会科学高等研究院-法国国家信息与自动化研究所）；PSL Research University（巴黎文理研究大学）；University of Toronto（多伦多大学）
Many types of distributional word embeddings (weakly) encode linguistic regularities as directions (the difference between "jump" and "jumped" will be in a similar direction to that of "walk" and "walked," and so on). Several attempts have been made to explain this fact. We respond to Allen and Hospedales' recent (ICML, 2019) theoretical explanation, which claims that word2vec and GloVe will encode linguistic regularities whenever a specific relation of paraphrase holds between the four words involved in the regularity. We demonstrate that the explanation does not go through: the paraphrase re…
University of Michigan（密歇根大学）；Ubiquiti, Inc.（优比快公司）
This paper presents a comparison of classification methods for linguistic typology for the purpose of expanding an extensive, but sparse language resource: the World Atlas of Language Structures (WALS) (Dryer and Haspelmath, 2013). We experimented with a variety of regression and nearest-neighbor methods for use in classification over a set of 325 languages and six syntactic rules drawn from WALS. To classify each rule, we consider the typological features of the other five rules; linguistic features extracted from a word-aligned Bible in each language; and genealogical features (genus and fam…
University of Copenhagen（哥本哈根大学）；University of Edinburgh（爱丁堡大学）
Character language models have access to surface morphological patterns, but it is not clear whether or how they learn abstract morphological regularities. We instrument a character language model with several probes, finding that it can develop a specific unit to identify word boundaries and, by extension, morpheme boundaries, which allows it to capture linguistic properties and regularities of these units. Our language model proves surprisingly good at identifying the selectional restrictions of English derivational morphemes, a task that requires both morphological and syntactic awareness. …
Institute of Computing Technology（计算技术研究所）；Chinese Academy of Sciences（中国科学院）
Recent work exhibited that distributed word representations are good at capturing linguistic regularities in language. This allows vector-oriented reasoning based on simple linear algebra between words. Since many different methods have been proposed for learning document representations, it is natural to ask whether there is also linear structure in these learned representations to allow similar reasoning at document level. To answer this question, we design a new document analogy task for testing the semantic regularities in document representations, and conduct empirical evaluations over se…
Johns Hopkins University（约翰斯·霍普金斯大学）；Université Paris Diderot（巴黎狄德罗大学）；CNRS（法国国家科学研究中心）；Microsoft Research AI（微软研究院人工智能部门）
Recurrent neural networks (RNNs) can learn continuous vector representations of symbolic structures such as sequences and sentences; these representations often exhibit linear regularities (analogies). Such regularities motivate our hypothesis that RNNs that show such regularities implicitly compile symbolic structures into tensor product representations (TPRs; Smolensky, 1990), which additively combine tensor products of vectors representing roles (e.g., sequence positions) and vectors representing fillers (e.g., particular words). To test this hypothesis, we introduce Tensor Product Decompos…
The paper presents a linguistic and computational model aiming at making the morphological structure of the lexicon emerge from the formal and semantic regularities of the words it contains. The model is word-based. The proposed morphological structure consists of (1) binary relations that connect each headword with words that are morphologically related, and especially with the members of its morphological family and its derivational series, and of (2) the analogies that hold between the words. The model has been tested on the lexicon of French using the TLFi machine readable dictionary.
This work has been motivated by two long term goals: to understand how humans learn language and to build programs that can understand language. Using a representation that makes the relevant features explicit is a prerequisite for successful learning and understanding. Therefore, I chose to represent relations between individual words explicitly in my model. Lexical attraction is defined as the likelihood of such relations. I introduce a new class of probabilistic language models named lexical attraction models which can represent long distance relations between words and I formalize this new…
Artificial Intelligence Research Institute（人工智能研究所）；SberDevices；HSE University（高等经济大学）
Linguistic analysis of language models is one of the ways to explain and describe their reasoning, weaknesses, and limitations. In the probing part of the model interpretability research, studies concern individual languages as well as individual linguistic structures. The question arises: are the detected regularities linguistically coherent, or on the contrary, do they dissonate at the typological scale? Moreover, the majority of studies address the inherent set of languages and linguistic structures, leaving the actual typological diversity knowledge out of scope. In this paper, we present …
University of Zurich（苏黎世大学）；Bar Ilan University（巴伊兰大学）；Allen Institute for Artificial Intelligence（艾伦人工智能研究所）
Historical linguists have identified regularities in the process of historic sound change. The comparative method utilizes those regularities to reconstruct proto-words based on observed forms in daughter languages. Can this process be efficiently automated? We address the task of proto-word reconstruction, in which the model is exposed to cognates in contemporary daughter languages, and has to predict the proto word in the ancestor language. We provide a novel dataset for this task, encompassing over 8,000 comparative entries, and show that neural sequence models outperform conventional metho…
Peking University（北京大学）；Beijing Institute of Big Data Research（北京大数据研究院）；Microsoft Research Asia（微软亚洲研究院）
Cross-lingual word embeddings aim to capture common linguistic regularities of different languages, which benefit various downstream tasks ranging from machine translation to transfer learning. Recently, it has been shown that these embeddings can be effectively learned by aligning two disjoint monolingual vector spaces through a linear transformation (word mapping). In this work, we focus on learning such a word mapping without any supervision signal. Most previous work of this task adopts parametric metrics to measure distribution differences, which typically requires a sophisticated alterna…
Beijing Normal University（北京师范大学）；Renmin University of China（中国人民大学）
Analogical reasoning is effective in capturing linguistic regularities. This paper proposes an analogical reasoning task on Chinese. After delving into Chinese lexical knowledge, we sketch 68 implicit morphological relations and 28 explicit semantic relations. A big and balanced dataset CA8 is then built for this task, including 17813 questions. Furthermore, we systematically explore the influences of vector representations, context features, and corpora on analogical reasoning. With the experiments, CA8 is proved to be a reliable benchmark for evaluating Chinese word embeddings.
University of Illinois at Urbana Champaign（伊利诺伊大学厄巴纳-香槟分校）
Real-valued word representations have transformed NLP applications; popular examples are word2vec and GloVe, recognized for their ability to capture linguistic regularities. In this paper, we demonstrate a {\em very simple}, and yet counter-intuitive, postprocessing technique -- eliminate the common mean vector and a few top dominating directions from the word vectors -- that renders off-the-shelf representations {\em even stronger}. The postprocessing is empirically validated on a variety of lexical-level intrinsic tasks (word similarity, concept categorization, word analogy) and sentence-lev…
Language Technologies Institute（语言技术研究所）；Carnegie Mellon University（卡内基梅隆大学）
We propose a new method for learning word representations using hierarchical regularization in sparse coding inspired by the linguistic study of word meanings. We show an efficient learning algorithm based on stochastic proximal methods that is significantly faster than previous approaches, making it possible to perform hierarchical sparse coding on a corpus of billions of word tokens. Experiments on various benchmark tasks---word similarity ranking, analogies, sentence completion, and sentiment analysis---demonstrate that the method outperforms or is competitive with state-of-the-art methods.…
Institute for Scientific Interchange (ISI)（科学交换研究所（ISI））；Universitat Politecnica de Catalunya（加泰罗尼亚理工大学）；CNR-ISC（意大利国家研究委员会-复杂系统研究所）；“Sapienza” Università di Roma（罗马第一大学（萨皮恩扎大学））
Human languages evolve continuously, and a puzzling problem is how to reconcile the apparent robustness of most of the deep linguistic structures we use with the evidence that they undergo possibly slow, yet ceaseless, changes. Is the state in which we observe languages today closer to what would be a dynamical attractor with statistically stationary properties or rather closer to a non-steady state slowly evolving in time? Here we address this question in the framework of the emergence of shared linguistic categories in a population of individuals interacting through language games. The obser…
The speech code is a vehicle of language: it defines a set of forms used by a community to carry information. Such a code is necessary to support the linguistic interactions that allow humans to communicate. How then may a speech code be formed prior to the existence of linguistic interactions? Moreover, the human speech code is discrete and compositional, shared by all the individuals of a community but different across communities, and phoneme inventories are characterized by statistical regularities. How can a speech code with these properties form? We try to approach these questions in the…
Text normalization techniques based on rules, lexicons or supervised training requiring large corpora are not scalable nor domain interchangeable, and this makes them unsuitable for normalizing user-generated content (UGC). Current tools available for Brazilian Portuguese make use of such techniques. In this work we propose a technique based on distributed representation of words (or word embeddings). It generates continuous numeric vectors of high-dimensionality to represent words. The vectors explicitly encode many linguistic regularities and patterns, as well as syntactic and semantic word …
Turbulence-like 5/3 spectral scaling in contextual representations of language as a complex system
College of Engineering, Peking University（北京大学工学院）；Ningbo Institute of Digital Twin, Eastern Institute of Technology（东方理工高等研究院宁波数字孪生研究院）；Shenzhen Tenfong Technology Co., Ltd.（深圳腾方科技有限公司）；Mechanical Engineering, Pennsylvania State University（宾夕法尼亚州立大学机械工程系）
Georgetown University（乔治城大学）；University of Washington（华盛顿大学）
Word embeddings learn implicit biases from linguistic regularities captured by word co-occurrence statistics. By extending methods that quantify human-like biases in word embeddings, we introduceValNorm, a novel intrinsic evaluation task and method to quantify the valence dimension of affect in human-rated word sets from social psychology. We apply ValNorm on static word embeddings from seven languages (Chinese, English, German, Polish, Portuguese, Spanish, and Turkish) and from historical English text spanning 200 years. ValNorm achieves consistently high accuracy in quantifying the valence o…
