Multiword Expressions in NLP

Computational Linguistics ◽

10.4018/978-1-4666-6042-7.ch010 ◽

2014 ◽

pp. 178-198 ◽

Cited By ~ 1

Author(s):

Alexander Gelbukh ◽

Olga Kolesnikova

Keyword(s):

Supervised Learning ◽

Language Processing ◽

Automatic Detection ◽

Research Problem ◽

Learning Methods ◽

Multiword Expressions ◽

Lexical Function

This chapter presents a survey of contemporary NLP research on Multiword Expressions (MWEs). MWEs pose a huge problem to precise language processing due to their idiosyncratic nature and diversity of their semantic, lexical, and syntactical properties. The chapter begins by considering MWEs definitions, describes some MWEs classes, indicates problems MWEs generate in language applications and their possible solutions, presents methods of MWE encoding in dictionaries and their automatic detection in corpora. The chapter goes into more detail on a particular MWE class called Verb-Noun Constructions (VNCs). Due to their frequency in corpus and unique characteristics, VNCs present a research problem in their own right. Having outlined several approaches to VNC representation in lexicons, the chapter explains the formalism of Lexical Function as a possible VNC representation. Such representation may serve as a tool for VNCs automatic detection in a corpus. The latter is illustrated on Spanish material applying some supervised learning methods commonly used for NLP tasks.

Download Full-text

Inductive semi-supervised learning methods for natural language processing

10.3115/1614101.1614106 ◽

2006 ◽

Author(s):

Anoop Sarkar ◽

Gholamreza Haffari

Keyword(s):

Natural Language Processing ◽

Natural Language ◽

Supervised Learning ◽

Language Processing ◽

Learning Methods

Download Full-text

Discovery Model Based on Analogies for Teaching Computer Programming

Mathematics ◽

10.3390/math9121354 ◽

2021 ◽

Vol 9 (12) ◽

pp. 1354

Author(s):

Javier Alejandro Jiménez Toledo ◽

César A. Collazos ◽

Manuel Ortega

Keyword(s):

Supervised Learning ◽

Language Processing ◽

Linked Data ◽

Computer Programming ◽

Synthetic Data ◽

Linguistic Analysis ◽

Complex Activity ◽

Learning Methods ◽

Model Based ◽

Space Modeling

Teaching the fundamentals of computer programming in a first course (CS1) is a complex activity for the professor and is also a challenge for them. Nowadays, there are several teaching strategies for dealing with a CS1 at the university, one of which is the use of analogies to support the abstraction process that a student needs to carry for the appropriation of fundamental concepts. This article presents the results of applying a discovery model that allowed for the extraction of patterns, linguistic analysis, textual analytics, and linked data when using analogies for teaching the fundamental concepts of programming by professors in a CS1 in university programs that train software developers. For that reason, a discovery model based on machine learning and text mining was proposed using natural language processing techniques for semantic vector space modeling, distributional semantics, and the generation of synthetic data. The discovery process was carried out using nine supervised learning methods, three unsupervised learning methods, and one semi-supervised learning method involving linguistic analysis techniques, text analytics, and linked data. The main findings showed that professors include keywords, which are part of the technical computer terminology, in the form of verbs in the statement of the analogy and combine them in quantitative contexts with neutral or positive phrases, where numerical examples, cooking recipes, and games were the most used categories. Finally, a structure is proposed for the construction of analogies to teach programming concepts and this was validated by the professors and students.

Download Full-text

Analysing the supervised learning methods for prediction of healthcare data in cloud environment A Survey

International Journal of Computer Sciences and Engineering ◽

10.26438/ijcse/v6i3.434438 ◽

2018 ◽

Vol 6 (3) ◽

pp. 434-438

Author(s):

N.M. Annigeri ◽

◽

S. Shetty ◽

A.P. Patil ◽

◽

...

Keyword(s):

Supervised Learning ◽

Cloud Environment ◽

Learning Methods ◽

Healthcare Data

Download Full-text

Adversarial training for few-shot text classification

Intelligenza Artificiale ◽

10.3233/ia-200051 ◽

2021 ◽

Vol 14 (2) ◽

pp. 201-214

Author(s):

Danilo Croce ◽

Giuseppe Castellucci ◽

Roberto Basili

Keyword(s):

Supervised Learning ◽

Language Processing ◽

Reproducing Kernel ◽

Generative Adversarial Networks ◽

Training Material ◽

Semantic Classification ◽

Universal Sentence ◽

Kernel Hilbert Spaces ◽

Supervised Methods ◽

Low Dimensional

In recent years, Deep Learning methods have become very popular in classification tasks for Natural Language Processing (NLP); this is mainly due to their ability to reach high performances by relying on very simple input representations, i.e., raw tokens. One of the drawbacks of deep architectures is the large amount of annotated data required for an effective training. Usually, in Machine Learning this problem is mitigated by the usage of semi-supervised methods or, more recently, by using Transfer Learning, in the context of deep architectures. One recent promising method to enable semi-supervised learning in deep architectures has been formalized within Semi-Supervised Generative Adversarial Networks (SS-GANs) in the context of Computer Vision. In this paper, we adopt the SS-GAN framework to enable semi-supervised learning in the context of NLP. We demonstrate how an SS-GAN can boost the performances of simple architectures when operating in expressive low-dimensional embeddings; these are derived by combining the unsupervised approximation of linguistic Reproducing Kernel Hilbert Spaces and the so-called Universal Sentence Encoders. We experimentally evaluate the proposed approach over a semantic classification task, i.e., Question Classification, by considering different sizes of training material and different numbers of target classes. By applying such adversarial schema to a simple Multi-Layer Perceptron, a classifier trained over a subset derived from 1% of the original training material achieves 92% of accuracy. Moreover, when considering a complex classification schema, e.g., involving 50 classes, the proposed method outperforms state-of-the-art alternatives such as BERT.

Download Full-text

Learning emotional word embeddings for sentiment analysis

Journal of Intelligent & Fuzzy Systems ◽

10.3233/jifs-201993 ◽

2021 ◽

pp. 1-13

Author(s):

Qingtian Zeng ◽

Xishi Zhao ◽

Xiaohui Hu ◽

Hua Duan ◽

Zhongying Zhao ◽

...

Keyword(s):

Sentiment Analysis ◽

Language Processing ◽

State Of The Art ◽

Research Problem ◽

Emotional Word ◽

Classification Model ◽

Data Sets ◽

Word Embeddings ◽

Real World Data ◽

Text Documents

Word embeddings have been successfully applied in many natural language processing tasks due to its their effectiveness. However, the state-of-the-art algorithms for learning word representations from large amounts of text documents ignore emotional information, which is a significant research problem that must be addressed. To solve the above problem, we propose an emotional word embedding (EWE) model for sentiment analysis in this paper. This method first applies pre-trained word vectors to represent document features using two different linear weighting methods. Then, the resulting document vectors are input to a classification model and used to train a text sentiment classifier, which is based on a neural network. In this way, the emotional polarity of the text is propagated into the word vectors. The experimental results on three kinds of real-world data sets demonstrate that the proposed EWE model achieves superior performances on text sentiment prediction, text similarity calculation, and word emotional expression tasks compared to other state-of-the-art models.

Download Full-text

A Survey on Contrastive Self-Supervised Learning

Technologies ◽

10.3390/technologies9010002 ◽

2020 ◽

Vol 9 (1) ◽

pp. 2

Author(s):

Ashish Jaiswal ◽

Ashwin Ramesh Babu ◽

Mohammad Zaki Zadeh ◽

Debapriya Banerjee ◽

Fillia Makedon

Keyword(s):

Computer Vision ◽

Supervised Learning ◽

Language Processing ◽

Large Scale ◽

Performance Comparison ◽

Extensive Review ◽

Future Directions ◽

Dominant Component ◽

Supervised Methods ◽

The Cost

Self-supervised learning has gained popularity because of its ability to avoid the cost of annotating large-scale datasets. It is capable of adopting self-defined pseudolabels as supervision and use the learned representations for several downstream tasks. Specifically, contrastive learning has recently become a dominant component in self-supervised learning for computer vision, natural language processing (NLP), and other domains. It aims at embedding augmented versions of the same sample close to each other while trying to push away embeddings from different samples. This paper provides an extensive review of self-supervised methods that follow the contrastive approach. The work explains commonly used pretext tasks in a contrastive learning setup, followed by different architectures that have been proposed so far. Next, we present a performance comparison of different methods for multiple downstream tasks such as image classification, object detection, and action recognition. Finally, we conclude with the limitations of the current methods and the need for further techniques and future directions to make meaningful progress.

Download Full-text