利用Markov网络抽取复述增强机器译文自动评价方法

在机器译文自动评价中,匹配具有相同语义、不同表达方式的词或短语是其中一个很大的挑战。许多研究工作提出从双语平行语料或可比语料中抽取复述来增强机器译文和人工译文的匹配。然而双语平行语料或可比语料不仅构建成本高,而且对少数语言对难以大量获取。我们提出通过构建词的Markov网络,从目标语言的单语文本中抽取复述的方法,并利用该复述提高机器译文自动评价方法与人工评价方法的相关性。在WMT14 Metrics task上的实验结果表明,我们从单语文本中提取复述方法的性能与从双语平行语料中提取复述方法的性能具有很强的可比性。因此,该文提出的方法可在保证复述质量的同时,降低复述抽取的成本。
     

MACHINE LEARNING OF MORPHOSYNTACTIC STRUCTURE: LEMMATIZING UNKNOWN SLOVENE WORDS

Automatic lemmatization is a core application for many language processing tasks. In inflectionally rich languages, such as Slovene, assigning the correct lemma (base form) to each word in a running text is not trivial, since for instance, nouns inflect for number and case, with a complex configuration of endings and stem modifications. The problem is especially difficult for unknown words, since word-forms cannot be matched against a morphological lexicon. This paper discusses a machine learning approach to the automatic lemmatization of unknown words in Slovene texts. We decompose the problem of learning to perform lemmatization into two subproblems: learning to perform morphosyntactic tagging of words in a text, and learning to perform morphological analysis, which produces the lemma from the word-form given the correct morphosyntactic tag. A statistics-based trigram tagger is used to learn morphosyntactic tagging and a first-order decision list learning system is used to learn rules for morphological analysis. We train the tagger on a manually annotated corpus consisting of 100,000 running words. We train the analyzer on open-class inflecting Slovene words, namely nouns, adjectives, and main verbs, together being characterized by more than 400 different morphosyntactic tags. The training set for the analyzer consists of a morphological lexicon containing 15,000 lemmas. We evaluate the learned model on word lists extracted from a corpus of Slovene texts containing 500,000 words, and show that our morphological analysis module achieves 98.6% accuracy, while the combination of the tagger and analyzer is 92.0% accurate on unknown inflecting Slovene words.     

A logical representation of Arabic questions toward automatic passage extraction from the Web

With the expanding growth of Arabic electronic data on the web, extracting information, which is actually one of the major challenges of the question-answering, is essentially used for building corpus of documents. In fact, building a corpus is a research topic that is currently referred to among some other major themes of conferences, in natural language processing (NLP), such as, information retrieval (IR), question-answering (QA), automatic summary (AS), etc. Generally, a question-answering system provides various passages to answer the user questions. To make these passages truly informative, this system needs access to an underlying knowledge base; this requires the construction of a corpus. The aim of our research is to build an Arabic question-answering system. In addition, analyzing the question must be the first step. Next, it is essential to retrieve a passage from the web that can serve as an appropriate answer. In this paper, we propose a method to analysis the question and retrieve the passage answer in the Arabic language. For the question analysis, five factual question types are processed. Additionally, our purpose is to experiment with the generation of a logic representation from the declarative form of each question. Several studies, deal with the logic approaches in question-answering, are discussed in other languages than the Arabic language. This representation is very promising because it helps us later in the selection of a justifiable answer. The accuracy of questions that are correctly analyzed and translated into the logic form achieved 64%. And then, the results of passages of texts that are automatically generated achieved an 87% score for accuracy and a 98% score for c@1.     

Exploring and exploiting a historical corpus for Arabic

This paper presents a historical Arabic corpus named HAC. At this early embryonic stage of the project, we report about the design, the architecture and some of the experiments which we have conducted on HAC. The corpus, and accordingly the search results, will be represented using a primary XML exchange format. This will serve as an intermediate exchange tool within the project and will allow the user to process the results offline using some external tools. HAC is made up of Classical Arabic texts that cover 1600 years of language use; the Quranic text, Modern Standard Arabic texts, as well as a variety of monolingual Arabic dictionaries. The development of this historical corpus assists linguists and Arabic language learners to effectively explore, understand, and discover interesting knowledge hidden in millions of instances of language use. We used techniques from the field of natural language processing to process the data and a graph-based representation for the corpus. We provided researchers with an export facility to render further linguistic analysis possible.     

Semantic indexing of Arabic texts for information retrieval system

As part of information retrieval systems (IRS) and in the context of the use of ontologies for documents and queries indexing, we propose and evaluate in this paper the contribution of this approach applied to Arabic texts. To do this we indexed a corpus of Arabic text using Arabic WordNet. The disambiguation of words was performed by applying the Lesk algorithm. The results obtained by our experiment allowed us to deduct the contribution of this approach in IRS for Arabic texts.     

Within-word pronunciation variation modeling for Arabic ASRs: a direct data-driven approach

Pronunciation variation is a major obstacle in improving the performance of Arabic automatic continuous speech recognition systems. This phenomenon alters the pronunciation spelling of words beyond their listed forms in the pronunciation dictionary, leading to a number of out of vocabulary word forms. This paper presents a direct data-driven approach to model within-word pronunciation variations, in which the pronunciation variants are distilled from the training speech corpus. The proposed method consists of performing phoneme recognition, followed by a sequence alignment between the observation phonemes generated by the phoneme recognizer and the reference phonemes obtained from the pronunciation dictionary. The unique collected variants are then added to dictionary as well as to the language model. We started with a Baseline Arabic speech recognition system based on Sphinx3 engine. The Baseline system is based on a 5.4 hours speech corpus of modern standard Arabic broadcast news, with a pronunciation dictionary of 14,234 canonical pronunciations. The Baseline system achieves a word error rate of 13.39%. Our results show that while the expanded dictionary alone did not add appreciable improvements, the word error rate is significantly reduced by 2.22% when the variants are represented within the language model.     

Learning-based word spotting system for Arabic handwritten documents

The retrieval of information from scanned handwritten documents is becoming vital with the rapid increase of digitized documents, and word spotting systems have been developed to search for words within documents. These systems can be either template matching algorithms or learning based. This paper presents a coherent learning based Arabic handwritten word spotting system which can adapt to the nature of Arabic handwriting, which can have no clear boundaries between words. Consequently, the system recognizes Pieces of Arabic Words (PAWs), then re-constructs and spots words using language models. The proposed system produced promising result for Arabic handwritten word spotting when tested on the CENPARMI Arabic documents database.     

Arabic speech synthesis and diacritic recognition

Text-to-speech system (TTS), known also as speech synthesizer, is one of the important technology in the last years due to the expanding field of applications. Several works on speech synthesizer have been made on English and French, whereas many other languages, including Arabic, have been recently taken into consideration. The area of Arabic speech synthesis has not sufficient progress and it is still in its first stage with a low speech quality. In fact, speech synthesis systems face several problems (e.g. speech quality, articulatory effect, etc.). Different methods were proposed to solve these issues, such as the use of large and different unit sizes. This method is mainly implemented with the concatenative approach to improve the speech quality and several works have proved its effectiveness. This paper presents an efficient Arabic TTS system based on statistical parametric approach and non-uniform units speech synthesis. Our system includes a diacritization engine. Modern Arabic text is written without mention the vowels, called also diacritic marks. Unfortunately, these marks are very important to define the right pronunciation of the text which explains the incorporation of the diacritization engine to our system. In this work, we propose a simple approach based on deep neural networks. Deep neural networks are trained to directly predict the diacritic marks and to predict the spectral and prosodic parameters. Furthermore, we propose a new simple stacked neural network approach to improve the accuracy of the acoustic models. Experimental results show that our diacritization system allows the generation of full diacritized text with high precision and our synthesis system produces high-quality speech.     

Towards Incremental Parsing of Natural Language Using Recursive Neural Networks

In this paper we develop novel algorithmic ideas for building a natural language parser grounded upon the hypothesis of incrementality. Although widely accepted and experimentally supported under a cognitive perspective as a model of the human parser, the incrementality assumption has never been exploited for building automatic parsers of unconstrained real texts. The essentials of the hypothesis are that words are processed in a left-to-right fashion, and the syntactic structure is kept totally connected at each step.Our proposal relies on a machine learning technique for predicting the correctness of partial syntactic structures that are built during the parsing process. A recursive neural network architecture is employed for computing predictions after a training phase on examples drawn from a corpus of parsed sentences, the Penn Treebank. Our results indicate the viability of the approach and lay out the premises for a novel generation of algorithms for natural language processing which more closely model human parsing. These algorithms may prove very useful in the development of efficient parsers.     

A text summarizer for Arabic

Automatic text summarization is an essential tool in this era of information overloading. In this paper we present an automatic extractive Arabic text summarization system where the user can cap the size of the final summary. It is a direct system where no machine learning is involved. We use a two pass algorithm where in pass one, we produce a primary summary using Rhetorical Structure Theory (RST); this is followed by the second pass where we assign a score to each of the sentences in the primary summary. These scores will help us in generating the final summary. For the final output, sentences are selected with an objective of maximizing the overall score of the summary whose size should not exceed the user selected limit. We used Rouge to evaluate our system generated summaries of various lengths against those done by a (human) news editorial professional. Experiments on sample texts show our system to outperform some of the existing Arabic summarization systems including those that require machine learning.     

Phonetically rich and balanced text and speech corpora for Arabic language

This paper describes the preparation, recording, analyzing, and evaluation of a new speech corpus for Modern Standard Arabic (MSA). The speech corpus contains a total of 415 sentences recorded by 40 (20 male and 20 female) Arabic native speakers from 11 different Arab countries representing three major regions (Levant, Gulf, and Africa). Three hundred and sixty seven sentences are considered as phonetically rich and balanced, which are used for training Arabic Automatic Speech Recognition (ASR) systems. The rich characteristic is in the sense that it must contain all phonemes of Arabic language, whereas the balanced characteristic is in the sense that it must preserve the phonetic distribution of Arabic language. The remaining 48 sentences are created for testing purposes, which are mostly foreign to the training sentences and there are hardly any similarities in words. In order to evaluate the speech corpus, Arabic ASR systems were developed using the Carnegie Mellon University (CMU) Sphinx 3 tools at both training and testing/decoding levels. The speech engine uses 3-emitting state Hidden Markov Models (HMM) for tri-phone based acoustic models. Based on experimental analysis of about 8?h of training speech data, the acoustic model is best using continuous observation’s probability model of 16 Gaussian mixture distributions and the state distributions were tied to 500 senones. The language model contains uni-grams, bi-grams, and tri-grams. For same speakers with different sentences, Arabic ASR systems obtained average Word Error Rate (WER) of 9.70%. For different speakers with same sentences, Arabic ASR systems obtained average WER of 4.58%, whereas for different speakers with different sentences, Arabic ASR systems obtained average WER of 12.39%.     

Extraction of terms and semantic relationships from Arabic texts for automatic construction of an ontology

The task of building an ontology from a textual corpus starts with the conceptualization phase, which extracts ontology concepts. These concepts are linked by semantic relationships. In this paper, we describe an approach to the construction of an ontology from an Arabic textual corpus, starting first with the collection and preparation of the corpus through normalization, removing stop words and stemming; then, to extract terms of our ontology, a statistical method for extracting simple and complex terms, called “the repeated segments method” are applied. To select segments with sufficient weight we apply the weighting method term frequency–inverse document frequency (TF–IDF), and to link these terms by semantic relationships we apply an automatic method of learning linguistic markers from text. This method requires a dataset of relationship pairs, which are extracted from two external resources: an Arabic dictionary of synonyms and antonyms and the lexical database Arabic WordNet. Finally, we present the results of our experimentation using our textual corpus. The evaluation of our approach shows encouraging results in terms of recall and precision.     

Investigation and modeling of the structure of texting language

Language usage over computer mediated discourses, such as chats, emails and SMS texts, significantly differs from the standard form of the language and is referred to as texting language (TL). The presence of intentional misspellings significantly decrease the accuracy of existing spell checking techniques for TL words. In this work, we formally investigate the nature and type of compressions used in SMS texts, and develop a Hidden Markov Model based word-model for TL. The model parameters have been estimated through standard machine learning techniques from a word-aligned SMS and standard English parallel corpus. The accuracy of the model in correcting TL words is 57.7%, which is almost a threefold improvement over the performance of Aspell. The use of simple bigram language model results in a 35% reduction of the relative word level error rates.     

A frame-based approach for capturing semantics from Arabic text for text-to-sign language MT

This paper describes the design and implementation of a computational model for Arabic natural language semantics, a semantic parser for capturing the deep semantic representation of Arabic text. The parser represents a major part of an Interlingua-based machine translation system for translating Arabic text into Sign Language. The parser follows a frame-based analysis to capture the overall meaning of Arabic text into a formal representation suitable for NLP applications that need for deep semantics representation, such as language generation and machine translation. We will show the representational power of this theory for the semantic analysis of texts in Arabic, a language which differs substantially from English in several ways. We will also show that the integration of WordNet and FrameNet in a single unified knowledge resource can improve disambiguation accuracy. Furthermore, we will propose a rule based algorithm to generate an equivalent Arabic FrameNet, using a lexical resource alignment of FrameNet1.3 LUs and WordNet3.0 synsets for English Language. A pilot study of motion and location verbs was carried out in order to test our system. Our corpus is made up of more than 2000 Arabic sentences in the domain of motion events collected from Algerian first level educational Arabic books and other relevant Arabic corpora.     

Finding an appropriate lexical diversity measurement for a small-sized corpus and its application to a comparative study of L2 learners’ writings

The present study investigates four kinds of lexical diversity measurement and a computational experiment with corpus processing and statistical test has been conducted to find out the most effective lexical diversity measurement in evaluating a small-sized corpus of 350?~?550 words. The results show that the D-estimate is the most appropriate among the four lexical diversity measurements which were compared in this research. Also the D-estimate showed more stable results than other measurements when the number of words varied between texts. The D-estimate was applied to measure the morphological and grammatical diversities of L2 learners of the Korean language, and conduct a statistical test on whether the mother tongues of L2 learners affect the degree of acquisition of grammatical morphemes. The test shows that the native languages of L2 learners learning Korean did not seem to have a significant impact.     

Predictive Text Typing System for the Ukrainian Language

This paper investigates the intellectualization of text input using a system for accelerated input of texts into digital devices with a view to constructing a model of a corpus of the Ukrainian spoken language and a text typing system based on this model. Such a system uses a smaller number of commands to input letters and predicts variants of words on the basis of the corpus of words and word combinations for communication. It is experimentally shown that the input of texts using four and six command keys is rather efficient for the constructed corpus.