In this paper we focus on patterns of colexification (co-expressions of form-meaning mapping in the lexicon) as an aspect of lexicalsemantic organization, and use them to build large scale synset graphs across BabelNet's typologically diverse set of 499 world languages.We introduce and compare several approaches: monolingual and cross-lingual colexification graphs, popular distributional models, and fusion approaches.The models are evaluated against human judgments on a semantic similarity task for nine languages.Our strong empirical findings also point to the importance of universality of our graph synset embedding representations with no need for any language-specific adaptation when evaluated on the lexical similarity task.The insights of our exploratory investigation of largescale colexification graphs could inspire significant advances in NLP across languages, especially for tasks involving languages which lack dedicated lexical resources, and can benefit from language transfer from large shared cross-lingual semantic spaces.

Identifying deception in online conversations is an important problem. A core component in interpreting deception is understanding how people perceive sincerity. Given that the larger proportion of human communication is non-verbal, decoding perceived sincerity is difficult in media where the only signal is written text. We introduce psycholinguistic features of perceived sincerity for two domains, dating and online games. Our results further the understanding of how language, deception, and perception connect.

Negation and speculation are fundamental elements of language and logic being studied in a variety of fields, including philosophy, logic, and psychology, as well as linguistics and computational linguistics. They are core components in information fact-checking and other applications (Morante and Sporleder 2012). In fact, natural language abounds in negative expressions (i.e., to encode “An act of denial; a negative statement, doctrine, etc.; a refusal or contradiction; a denial of something.” (OED 1989)) and speculative devices (i.e., to encode “a conclusion, opinion, view, or series of these, reached by abstract or hypothetical reasoning” (OED 1989)) at different linguistic levels (i.e., lexical, morphological and syntactic, semantic, pragmatic, and prosodic). Negation and modality (a category of linguistic meaning underlying the expression of possibility and necessity (Fintel 2006)) are two of the most discussed phenomena within linguistics, and more recently in natural language processing (NLP) (Morante and Sporleder 2012). Negation and speculation (usually encoded through modal devices) are particularly challenging for NLP not only because they add additional meaning and structure to process, but also because there are many different ways to express such linguistic phenomena.A common task in NLP is to extract, infer, and make sense of facts from the world around us—that is, to uncover factual information from text. Popular applications like information extraction and text mining have attracted considerable attention in and beyond the community with the purpose of populating large ontologies and using them for fact-checking (Cohen, Hamilton, and Turner 2011). Within this context, speculative language plays a particular role as it reflects the attitude of the writer/speaker toward the truth value of certain facts.This short book by Cruz Diáz and Maña López is a nice introduction to negation and speculation in linguistics and natural language processing, covering a wide range of key aspects from theory (definition and classification types) to practice (annotated corpora and automatic models of negation and speculation detection). The book is targeted toward students of computational linguistics and researchers with adequate background in NLP and machine learning as well as software engineers interested in available resources for research and development. The authors do a good job addressing the main issues, thus filling an important gap in the computational linguistics field: the lack of surveys and literature review on various aspects of language and language processing. As in many academic disciplines, evaluation and synthesis of prior research have become imperative in keeping up with the exponentially growing scientific literature. To this we should also add new technology and trends hitting the language technology market today. Such surveys are crucial in assisting students, academics, researchers, and end-users in quickly finding, evaluating, and synthesizing a broad content landscape encapsulating all the essential developments on the topic.The book consists of six chapters, each chapter addressing a single specific topic. Although the specific chapters on negation and speculation are (almost) self-contained, the rest of the book is logically connected, through applications, resources, and future trends. Chapters 2–5 follow a similar structure, with content relevant to the chapter’s topic, ending with two subsections on summary and conclusions, and future reading and relevant resources. In Chapter 1 the authors briefly introduce the linguistic phenomena of negation and speculation, drawing attention to their importance as well as to inherent challenges in linguistics and NLP. They also overview applications that can benefit from the detection of such linguistic devices.Chapters 2 and 3 form the core sections, covering about half of the book’s content. Both chapters introduce a few linguistic taxonomies of various types of negation and speculation with some reference to relevant linguistic literature. These theoretical aspects are then followed by a comprehensive review of computational linguistics works focusing on models, corpora, and other resources used in the automatic detection of negation and speculation. Although the reader would appreciate the briefly introduced computational approaches according to various genres, languages, and types of modeling (from rule-based to machine learning and hybrid approaches), the chapters provide only a short reference to specific research work without much comparative analysis of the models discussed. Such comparison is necessary to understand which models work better, how they do so, and in what settings. Moreover, the authors do not provide a clear connec- tion between the linguistic classes of negation and speculation and the computational models of these devices. There is no doubt that newcomers to these topics will appreciate such a comprehensive survey of computational approaches. However, it would have bene- fited the reader if a case study or detailed analysis was included for negation and spec- ulation with a discussion of the model, data, experiments, and comprehensive results.In the second part of the book, which includes Chapters 4 and 5, the authors discuss various tools and applications that can benefit from negation and speculation detection, and summarize a set of relevant resources for those interested in building them. Whereas the subsections on automatic detection are comprehensive and informative, those on chapter summary and conclusions, and future reading and relevant resources, are very short and uninspiring. The book ends with a discussion of the key challenges in this domain followed by some trends and possible future developments.In summary, the book should be considered a quick reference guide to research on automatic detection of negation and speculation in language, as it provides a valuable introduction to these topics, covering relevant theory, computational models, applications, and resources. However, given the effort involved in writing research surveys and the impact they should have on the community, this book should be a reference point for the scholar who wants to get a flash overview of the covered topic in linguistics and computational linguistics, as well as a good starting point for future discussions in NLP on potential developments. This is one of the aspects this book does not address well.First, the definition of negation used here is limited to a clearly demarcated linguistic phenomenon whose purpose is to reverse the polarity of a sentence or its parts (i.e., sentential negation not, as in “A is NOT B” constructions). Indeed, in the logical tradition, as Givón (1989) noted, negation is simply defined as a truth-value operator. However, he argues that, when it comes to language, negation is much more than that, and that a real understanding of the concept must take contextual factors into account (Givón 1989). More recent research in linguistics has shown that the nature of negation finds itself at the crossroads between truth-value operator and speech act (e.g., the speaker adds to the utterance a belief or attitude toward certain phenomena of reality), where the purpose of negation is often to produce meanings beyond the one of a reversed affirmation “it is not the case that X” (Roitman and Birkelund 2017). Additionally, because most of the research in linguistics and computational linguistics has focused on the negative form not and the like, we do not know much about how other expressions of negation are processed, understood, and used computationally. The authors are right in saying that negation and speculation detection are popular yet still emerging topics in NLP. However, it would have been nice to see a more in-depth discussion on possible ways the community can and should extend the scope of research beyond the mere presence or absence of negation and speculation cue words or phrases.Second, the book does not cover relevant and important applications, domains, or tasks like dialogue systems and fact-checking (i.e., truth and deception detection), which have received considerable attention in recent years not only in academia but also in the industry (Chen et al. 2017; Thorne and Vlachos 2018; Vosoughi, Roy, and Aral 2018). The narrow definitions of negation and speculation as well as the many computational detection models introduced here might or might not apply to these application domains. Most of the linguistic and computational examples presented in this book rely on research articles and biomedical text as data, yet there is minimal discussion about the implications of these theories and models to other genres like dialogues, social media, and legal text.In summary, this book is a good reference for most works on negation and speculation detection in linguistics and computational linguistics, and, hopefully, a starting point for discussion in the NLP community. This is an important area of research and I look forward to the exciting breakthroughs it will bring.

Understanding how people perceive sincerity is important in identifying malevolent deceptive behavior in social media. This is a difficult problem because the only signal in most such cases is written text, and the larger proportion of human communication is non-verbal. Furthermore, perception is very subjective, and people adapt to each other's perceptions. We define the problem of decoding sincerity perception for two domains, dating and online games. We introduce and evaluate new psycholinguistic features of perceived sincerity. Our results shed light into the connection between language, deception, and perception, and underline the challenges and difficulty of assessing perceptions from written text.

There are users who generate significant amounts of domain knowledge in online forums or community question and answer (CQA) websites. Existing literature defines them as `experts.' These users attain such statuses by providing multiple relevant answers to the question askers. Past works have focused on recommending relevant posts to these users. With the rise of web forums where certified experts answer questions, strategies that are tailored towards addressing the new type of experts will be beneficial. In this paper, we identify a new type of user called `designated experts' (i.e., users designated as domain experts by the web administrators). These are the experts who are guaranteed by web administrators to be an expert in a given domain. Our focus is on how we can capture the unique behavior of designated experts in an online domain. We have noticed designated experts have different behaviors compared to CQA experts. In particular, unlike existing CQAs, only one designated expert responds to any given thread. To capture this intuition, we introduce a matrix factorization algorithm with regularization to capture the behavior. Our results show that the regularization method improves the performance significantly compared to the baseline approach.

Recognition of causality is important to achieve natural language discourse under-standing. Previous approaches rely on shallow linguistic features. In this work, we propose to identify causality in verb-noun pairs by exploiting deeper seman-tics of nouns and verbs. Particularly, we acquire and employ three novel types of knowledge: (1) semantic classes of nouns with a high and low tendency to encode causality along with information regard-ing metonymies, (2) data-driven seman-tic classes of verbal events with the least tendency to encode causality, and (3) ten-dencies of verb frames to encode causal-ity. Using these knowledge sources, we achieve around 15 % improvement in F-score over a supervised classifier trained using linguistic features. 1

Abstract Reciprocity is a pervasive concept that plays an important role in governing people's behavior, judgments, and thus their social interactions. In this paper we present an analysis of the concept of reciprocity as expressed in English and a way to model it. At a larger structural level the reciprocity model will induce representations and clusters of relations between interpersonal verbs. In particular, we introduce an algorithm that semi-automatically discovers patterns encoding reciprocity based on a set of simple yet effective pronoun templates. Using the most frequently occurring patterns we queried the web and extracted 13,443 reciprocal instances, which represent a broad-coverage resource. Unsupervised clustering procedures are performed to generate meaningful semantic clusters of reciprocal instances. We also present several extensions (along with observations) to these models that incorporate meta-attributes like the verbs' affective value, identify gender differences between participants, consider the textual context of the instances, and automatically discover verbs with certain presuppositions. The pattern discovery procedure yields an accuracy of 97 per cent, while the clustering procedures – clustering with pairwise membership and clustering with transitions – indicate accuracies of 91 per cent and 64 per cent, respectively. Our affective value clustering can predict an unknown verb's affective value (positive, negative, or neutral) with 51 per cent accuracy, while it can discriminate between positive and negative values with 68 per cent accuracy. The presupposition discovery procedure yields an accuracy of 97 per cent.

This paper presents the Topic-Aspect Model (TAM), a Bayesian mixture model which jointly discovers topics and aspects. We broadly define an aspect of a document as a characteristic that spans the document, such as an underlying theme or perspective. Unlike previous models which cluster words by topic or aspect, our model can generate token assignments in both of these dimensions, rather than assuming words come from only one of two orthogonal models. We present two applications of the model. First, we model a corpus of computational linguistics abstracts, and find that the scientific topics identified in the data tend to include both a computational aspect and a linguistic aspect. For example, the computational aspect of GRAMMAR emphasizes parsing, whereas the linguistic aspect focuses on formal languages. Secondly, we show that the model can capture different viewpoints on a variety of topics in a corpus of editorials about the Israeli-Palestinian conflict. We show both qualitative and quantitative improvements in TAM over two other state-of-the-art topic models.

In this article we explore the syntactic and semantic properties of prepositions in the context of the semantic interpretation of nominal phrases and compounds. We investigate the problem based on cross-linguistic evidence from a set of six languages: English, Spanish, Italian, French, Portuguese, and Romanian. The focus on English and Romance languages is well motivated. Most of the time, English nominal phrases and compounds translate into constructions of the form N P N in Romance languages, where the P (preposition) may vary in ways that correlate with the semantics. Thus, we present empirical observations on the distribution of nominal phrases and compounds and the distribution of their meanings on two different corpora, based on two state-of-the-art classification tag sets: Lauer's set of eight prepositions and our list of 22 semantic relations. A mapping between the two tag sets is also provided. Furthermore, given a training set of English nominal phrases and compounds along with their translations in the five Romance languages, our algorithm automatically learns classification rules and applies them to unseen test instances for semantic interpretation. Experimental results are compared against two state-of-the-art models reported in the literature.

For readers of traditional textbooks such as that of Partee, ter Meulen, and Wall (1990), the term 'mathematical linguistics' denotes a rather narrowly circumscribed set of issues including automata theory, set theory, and lambda calculus, with maybe a little formal language theory thrown in.Kornai's contribution is refreshingly different in that he treats, in this relatively compact volume, practically all areas of linguistics, phonetics, and speech and language processing.Kornai's motivation for writing this book is to present "a single entry point to the central methods and concepts of linguistics that are made largely inaccessible to the mathematician, computer scientist, or engineer by the surprisingly adversarial style of argumentation . . .and the proliferation of unmotivated notation and formalism . . .all too often encountered in research papers and monographs in the humanities" (page viii).There is no question that much of what passes for rigor (mathematical and scientific) in linguistics is a joke and that there is clearly a need for any work that can place the field on a more solid footing.It also seems likely that Kornai is the only person who could have written this book.The book is divided into ten chapters, including a short introductory chapter, which lays the groundwork and identifies the potential audience, and a concluding chapter where Kornai reveals his own views on what is important in the field, which in the interests of balance he has largely suppressed throughout the book.Chapter 2 is also introductory in that it presents basic concepts of generation (via a ruleset), axioms, and string rewriting.The main chapters (3-9) deal with a variety of topic areas relating to language and speech, starting with phonology in Chapter 3.This chapter introduces the notion of phonemes, distinctive features, autosegmental phonology, and computation using finite automata.Kornai offers many details that are of course lacking in most linguistic treatments, such as a proof that the number of well-formed association lines between two tiers of length n is asymptotically (6 + 4 √ 2) n .Chapter 4 deals with morphology, which for Kornai includes not only word formation, but also prosody (including stress assignment and moraic structure), as well as Optimality Theory and Zipf's law.The fifth chapter treats syntax, including categorial grammar, phrase structure, dependency frameworks, valency, and weighted models of grammar, ending with a discussion of weighted finite automata and hidden Markov models.In the context of weighted models, Kornai implies that Chomsky's original notion of degree of grammaticality fits naturally as an instance of a weighted model with a particular semiring; of course, exactly what the ⊕ and ⊗ operators of that semiring map to remain to

The acquisition of semantic knowledge is paramount for any application that requires a deep understanding of natural language text. Motivated by the problem of building a noun phrase-level semantic parser and adapting it to various applications, such as machine translation and multilingual question answering, in this paper we present a domain-independent model for noun phrase semantic interpretation. We investigate the problem based on cross-linguistic evidence from a set of four Romance languages: Spanish, Italian, French, and Romanian. The focus on Romance languages is well motivated. It is generally the case that English noun phrases translate into constructions of the form "N P N" in Romance languages where, as we will show, the P (preposition) varies in ways that correlate with the semantics. Thus, based on a set of 22 semantic interpretation categories (such as PART-WHOLE, AGENT, POSSESSION) we present empirical observations regarding the distribution of these semantic categories in a cross-lingual corpus and their mapping to various syntactic constructions in English and Romance. Furthermore, given a training set of English noun phrases along with their translations in the four Romance languages, our algorithm automatically learns classification rules and applies them to unseen noun phrase instances for semantic interpretation. Experimental results are compared against a state-of-the-art model reported in the literature.

An important problem in knowledge discovery from text is the automatic extraction of semantic relations. This paper presents a supervised, semantically intensive, domain independent approach for the automatic detection of part-whole relations in text. First an algorithm is described that identifies lexico-syntactic patterns that encode part-whole relations. A difficulty is that these patterns also encode other semantic relations, and a learning method is necessary to discriminate whether or not a pattern contains a part-whole relation. A large set of training examples have been annotated and fed into a specialized learning system that learns classification rules. The rules are learned through an iterative semantic specialization (ISS) method applied to noun phrase constituents. Classification rules have been generated this way for different patterns such as genitives, noun compounds, and noun phrases containing prepositional phrases to extract part-whole relations from them. The applicability of these rules has been tested on a test corpus obtaining an overall average precision of 80.95% and recall of 75.91%. The results demonstrate the importance of word sense disambiguation for this task. They also demonstrate that different lexico-syntactic patterns encode different semantic information and should be treated separately in the sense that different clarification rules apply to different patterns.

This paper presents an open-domain textual Question-Answering system that uses several feedback loops to enhance its performance. These feedback loops combine in a new way statistical results with syntactic, semantic or pragmatic information derived from texts and lexical databases. The paper presents the contribution of each feedback loop to the overall performance of 76% human-assessed precise answers.

This paper presents the architecture, operation and results obtained with the LASSO Question Answering system developed in the Natural Language Processing Laboratory at SMU. To find answers, the system relies on a combination of syntactic and semantic techniques. The search for the answer is based on a novel form of indexing called paragraph indexing. A score of 55.5% for short answers and 64.5% for long answers was achieved at the TREC-8 competition.