Large Web Corpus Research Articles

Language contact between Italian and English: a case study on nouns ending in the suffix -ing

Abstract The article deals with how English deverbal nouns with the suffix -ing have been imported into Italian. The focus is on the semantic characteristics of these borrowed nouns in Italian and, in particular, on the question of whether they have been borrowed not as simple sign-concept pairings but with argument and event structure. In previous research, it has been claimed that argument and event structure need to be licensed by some overt functional element. Hence, borrowed deverbal nouns should have argument structure and event structure only if they have an overt affix, in other words, only if the forms are not borrowed holistically as unsegmented words but retain internal morphological structure, implying that morphological borrowing of the affix has occurred. When a foreign affix combines with native bases of the recipient language, this is often considered an important criterion for morphological borrowing, which is clearly not the case for the suffix -ing in Italian. Here, it will be shown that contrary to expectation, numerous occurrences with argument and event structure may be found in a large Italian web corpus for a sample of English deverbal nouns ending in -ing, borrowed into (certain registers) of Italian.

Die Komparativ- und Superlativformen besser/am besten/das Beste als Varianten in modalen Vergleichskonstruktionen

It is well known that linguistic variants play a key role in the acquisition of language skills in the first, second or foreign language as well as in writing and translation processes and in general in communicative interactions. Thus, a major research goal is the systematic investigation of intra- and interlinguistic variation. Due to its complexity, its qualitative-quantitative analysis continues to be a challenging issue, but it seems to become more and more feasible thanks to both the possibility of compiling very large corpora and the availability of high-performing corpus-linguistic tools.
 The paper discusses a corpus-linguistic pilot study concerning the use of besser, am besten and das Beste as pragmatic markers in a cross-linguistic perspective. In particular, the analysis focusses on selected superlative and comparative constructions on the left periphery used for expressing advice. The data basis consists mainly of German and Italian comparable very large web corpora and, to a lesser extent, of bilingual sentence pairs drawn from parallel corpora. As will be shown, even restricting the analysis to a very small segment of microvariation, in both languages the modal constructions appear to be characterized by the combination of numerous overlapping and interplaying variants and by different tendencies in language use.

This! Identifying New Sentiment Slang Through Orthographic Pleonasm Online: Yasss Slay Gorg Queen Ilysm

Identifying neologisms is important for natural language processing of social web text when informal language is standard and youth slang is common. For example, failing to identify neologisms can reduce the accuracy of the lexical sentiment analysis if opinions are frequently expressed in words that are too new to be in the sentiment dictionary. This article proposes a method based on orthographic pleonasm to identify emotion-related neologisms in the social web: finding words with the most different letter repetition spelling variations. For this method, nondictionary words are extracted from a large social web corpus, spelling standardization is applied, and then words are ranked in a decreasing order of spelling variation frequency. Words with the most spelling variations are then keyword-in-context-analyzed for semantic context. Applied to a collection of comments on YouTube influencers, this method founds neologisms such as slay and early as positive terms, mixed with traditional sentiment words, exclamations, and nouns. Although orthographic pleonasm was originally used to express the speaker's rhythm and tone of voice, it is also used for initialisms in a way that is difficult to vocalize. The method is, therefore, a practical way to identify new sentiment slang, including both normal words and initialisms.

When complementation gets specific: A study of collocational preferences in verb–object combinations in Norwegian

AbstractThis study investigates to what extent there are collocational preferences in the verb–object combinations of a large corpus of Norwegian and how important recurrent combinations are in usage. The material has been extracted from a large web-corpus of 700 million tokens and consists of dependency-based verb–object combinations. The overall importance of collocational preferences is demonstrated by the fact that the most frequent 5% of the verb–object combinations account for as much as 64% of the verb–object tokens in the material. The database of verb–object combinations contains measures of collocational strength and thereby allows us to model the mutual strength between the exemplars in the clusters found with individual verbs. Based on some studies of individual verbs and verb pairs, it seems safe to assume that speakers do distinguish between and prefer certain conventional verb–object combinations to other equally grammatical, equally transparent and equally understandable alternatives, and that speakers have access to complementation information at the level of exemplars.

Twee vroege constructiegrammaticale studies in Nederlandse Taalkunde*

Abstract Two early construction-grammatical studies in Dutch Linguistics. Looking back on ) and This article looks back on the articles by and , on the double object construction and a Dutch equivalent of the English ‘way’-construction, respectively, which can be considered the first two articles to have appeared in Nederlandse Taalkunde that address problems of Dutch grammar from an explicitly constructionist perspective. I illustrate how, on the basis of data from a large web corpus such as NLCOW14, the formal and semantic analyses offered in the two articles can be refined in a number of ways, but I also show that they include hypotheses and ideas that are still most relevant and relate to topical discussions in construction grammar.

Stripping paradigmatic relations out of the syntax

This contribution aims at showing how paradigms and associative relations can be integrated into word-formation, with special attention paid to compounding. In this regard, we will take into account a phenomenon at the border between derivation and compounding, namely formations like susherzig ‘sweet-hearted’, in which -ig is an adjective-forming suffix and AN a possible compound. To do so, we will explore data available from a large web corpus, on whose basis we will show how syntagmatic and paradigmatic relations developed in syntax find their way into word formation. We will show that the most productive compounding schemas as they are currently assumed in Construction Morphology give rise to processes of semi-affixation which are a first step toward derivation proper.

Functional Text Dimensions for the annotation of web corpora

This paper presents an approach to classifying large web corpora into genres by means of Functional Text Dimensions (FTDs). This offers a topological approach to text typology in which the texts are described in terms of their similarity to prototype genres. The suggested set of categories is designed to be applicable to any text on the web and to be reliable in annotation practice. Interannotator agreement results show that the suggested categories produce Krippendorff's α at above 0.76. In addition to the functional space of eighteen dimensions, similarity between annotated documents can be described visually within a space of reduced dimensions obtained through t-distributed Statistical Neighbour Embedding. Reliably annotated texts also provide the basis for automatic genre classification, which can be done in each FTD, as well as as within the space of reduced dimensions. An example comparing texts from the Brown Corpus, the BNC and ukWac, a large web corpus, is provided.

How words behave in other languages

Abstract This paper undertakes a systematic investigation into the use of German Nazi vocabulary in English. Nazi vocabulary is checked for frequency of occurrence in a large web corpus of English and then, where it occurs, for reference to Nazi discourse. Next, its frequency is compared to similar French and German web corpora, showing whether or not the use of Nazi vocabulary outside German is unique to English and whether or not its current usage differs between German and the borrowing languages. Finally, the use of two words that occur with similar frequencies in all three languages – judenrein and Blitzkrieg – and of two words that occur there with the highest difference in frequency – Anschluss and Lebensraum – is investigated in detail by means of the Sketch Engine corpus tool, including an analysis of collocations which indicate contexts of usage. The results can inform further research into lexical borrowing by demonstrating that borrowed words may be used in ways that differ notably from their use in the donor language.

The framing of Muslims on the Spanish Internet

AbstractIn this work we study the representation of Muslims on the Internet in Spain. After the terrorist attacks in Europe, Islamophobia and Muslimophobia have grown considerably in our society. There is a strong rejection of Muslim groups and individuals, they are perceived not only as different, but also as dangerous and violent. We follow a cognitive linguistics approach using corpus linguistics as a methodology in order to know which concepts are related to Muslims in discourse. We have used three corpora: the Spanish part of the esTenTen corpus, which is a large web corpus intended to give a picture of the Spanish language on the Internet; a Twitter corpus encompassing tweets published by five main political parties in Spain and their candidates in 2015-2016; and a third corpus of articles on the topic “Muslims” from four important digital newspapers (

SENTIMENT CLASSIFICATION of MOVIE REVIEWS USING CONTEXTUAL VALENCE SHIFTERS

We present two methods for determining the sentiment expressed by a movie review. The semantic orientation of a review can be positive, negative, or neutral. We examine the effect of valence shifters on classifying the reviews. We examine three types of valence shifters: negations, intensifiers, and diminishers. Negations are used to reverse the semantic polarity of a particular term, while intensifiers and diminishers are used to increase and decrease, respectively, the degree to which a term is positive or negative. The first method classifies reviews based on the number of positive and negative terms they contain. We use the General Inquirer to identify positive and negative terms, as well as negation terms, intensifiers, and diminishers. We also use positive and negative terms from other sources, including a dictionary of synonym differences and a very large Web corpus. To compute corpus‐based semantic orientation values of terms, we use their association scores with a small group of positive and negative terms. We show that extending the term‐counting method with contextual valence shifters improves the accuracy of the classification. The second method uses a Machine Learning algorithm, Support Vector Machines. We start with unigram features and then add bigrams that consist of a valence shifter and another word. The accuracy of classification is very high, and the valence shifter bigrams slightly improve it. The features that contribute to the high accuracy are the words in the lists of positive and negative terms. Previous work focused on either the term‐counting method or the Machine Learning method. We show that combining the two methods achieves better results than either method alone.

The effects of fitness functions on genetic programming‐based ranking discovery for Web search

AbstractGenetic‐based evolutionary learning algorithms, such as genetic algorithms (GAs) and genetic programming (GP), have been applied to information retrieval (IR) since the 1980s. Recently, GP has been applied to a new IR task—discovery of ranking functions for Web search—and has achieved very promising results. However, in our prior research, only one fitness function has been used for GP‐based learning. It is unclear how other fitness functions may affect ranking function discovery for Web search, especially since it is well known that choosing a proper fitness function is very important for the effectiveness and efficiency of evolutionary algorithms. In this article, we report our experience in contrasting different fitness function designs on GP‐based learning using a very large Web corpus. Our results indicate that the design of fitness functions is instrumental in performance improvement. We also give recommendations on the design of fitness functions for genetic‐based information retrieval experiments.