Enfrentamento do problema das divergências de tradução por um sistema de tradução automática: um exercício exploratório

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

A interlíngua e o alinhamento léxico-conceitual da base de dados bilíngue REBECA

In the architecture of a natural language processing system based on linguistic knowledge, two types of component are important: the knowledge databases and the processing modules. One of the knowledge databases is the lexical database, which is responsible for providing the lexical unities and its properties to the processing modules. The systems that process two or more languages require bilingual and/or multilingual lexical databases. These databases can be constructed by aligning distinct monolingual databases. In this paper, we present the interlingua and the strategy of aligning the two monolingual databases in REBECA, which only stores concepts from the “wheeled vehicle” domain.

Extractive summarization using complex networks and syntactic dependency

The realization that statistical physics methods can be applied to analyze written texts represented as complex networks has led to several developments in natural language processing, including automatic summarization and evaluation of machine translation. Most importantly, so far only a few metrics of complex networks have been used and therefore there is ample opportunity to enhance the statistics-based methods as new measures of network topology and dynamics are created. In this paper, we employ for the first time the metrics betweenness, vulnerability and diversity to analyze written texts in Brazilian Portuguese. Using strategies based on diversity metrics, a better performance in automatic summarization is achieved in comparison to previous work employing complex networks. With an optimized method the Rouge score (an automatic evaluation method used in summarization) was 0.5089, which is the best value ever achieved for an extractive summarizer with statistical methods based on complex networks for Brazilian Portuguese. Furthermore, the diversity metric can detect keywords with high precision, which is why we believe it is suitable to produce good summaries. It is also shown that incorporating linguistic knowledge through a syntactic parser does enhance the performance of the automatic summarizers, as expected, but the increase in the Rouge score is only minor. These results reinforce the suitability of complex network methods for improving automatic summarizers in particular, and treating text in general. (C) 2011 Elsevier B.V. All rights reserved.

Automatic induction of lexical features

This thesis concerns artificially intelligent natural language processing systems that are capable of learning the properties of lexical items (properties like verbal valency or inflectional class membership) autonomously while they are fulfilling their tasks for which they have been deployed in the first place. Many of these tasks require a deep analysis of language input, which can be characterized as a mapping of utterances in a given input C to a set S of linguistically motivated structures with the help of linguistic information encoded in a grammar G and a lexicon L: G + L + C → S (1) The idea that underlies intelligent lexical acquisition systems is to modify this schematic formula in such a way that the system is able to exploit the information encoded in S to create a new, improved version of the lexicon: G + L + S → L' (2) Moreover, the thesis claims that a system can only be considered intelligent if it does not just make maximum usage of the learning opportunities in C, but if it is also able to revise falsely acquired lexical knowledge. So, one of the central elements in this work is the formulation of a couple of criteria for intelligent lexical acquisition systems subsumed under one paradigm: the Learn-Alpha design rule. The thesis describes the design and quality of a prototype for such a system, whose acquisition components have been developed from scratch and built on top of one of the state-of-the-art Head-driven Phrase Structure Grammar (HPSG) processing systems. The quality of this prototype is investigated in a series of experiments, in which the system is fed with extracts of a large English corpus. While the idea of using machine-readable language input to automatically acquire lexical knowledge is not new, we are not aware of a system that fulfills Learn-Alpha and is able to deal with large corpora. To instance four major challenges of constructing such a system, it should be mentioned that a) the high number of possible structural descriptions caused by highly underspeci ed lexical entries demands for a parser with a very effective ambiguity management system, b) the automatic construction of concise lexical entries out of a bulk of observed lexical facts requires a special technique of data alignment, c) the reliability of these entries depends on the system's decision on whether it has seen 'enough' input and d) general properties of language might render some lexical features indeterminable if the system tries to acquire them with a too high precision. The cornerstone of this dissertation is the motivation and development of a general theory of automatic lexical acquisition that is applicable to every language and independent of any particular theory of grammar or lexicon. This work is divided into five chapters. The introductory chapter first contrasts three different and mutually incompatible approaches to (artificial) lexical acquisition: cue-based queries, head-lexicalized probabilistic context free grammars and learning by unification. Then the postulation of the Learn-Alpha design rule is presented. The second chapter outlines the theory that underlies Learn-Alpha and exposes all the related notions and concepts required for a proper understanding of artificial lexical acquisition. Chapter 3 develops the prototyped acquisition method, called ANALYZE-LEARN-REDUCE, a framework which implements Learn-Alpha. The fourth chapter presents the design and results of a bootstrapping experiment conducted on this prototype: lexeme detection, learning of verbal valency, categorization into nominal count/mass classes, selection of prepositions and sentential complements, among others. The thesis concludes with a review of the conclusions and motivation for further improvements as well as proposals for future research on the automatic induction of lexical features.

Deep learning for computer vision: a comparison between convolutional neural networks and hierarchical temporal memories on object recognition tasks

In recent years, Deep Learning techniques have shown to perform well on a large variety of problems both in Computer Vision and Natural Language Processing, reaching and often surpassing the state of the art on many tasks. The rise of deep learning is also revolutionizing the entire field of Machine Learning and Pattern Recognition pushing forward the concepts of automatic feature extraction and unsupervised learning in general. However, despite the strong success both in science and business, deep learning has its own limitations. It is often questioned if such techniques are only some kind of brute-force statistical approaches and if they can only work in the context of High Performance Computing with tons of data. Another important question is whether they are really biologically inspired, as claimed in certain cases, and if they can scale well in terms of "intelligence". The dissertation is focused on trying to answer these key questions in the context of Computer Vision and, in particular, Object Recognition, a task that has been heavily revolutionized by recent advances in the field. Practically speaking, these answers are based on an exhaustive comparison between two, very different, deep learning techniques on the aforementioned task: Convolutional Neural Network (CNN) and Hierarchical Temporal memory (HTM). They stand for two different approaches and points of view within the big hat of deep learning and are the best choices to understand and point out strengths and weaknesses of each of them. CNN is considered one of the most classic and powerful supervised methods used today in machine learning and pattern recognition, especially in object recognition. CNNs are well received and accepted by the scientific community and are already deployed in large corporation like Google and Facebook for solving face recognition and image auto-tagging problems. HTM, on the other hand, is known as a new emerging paradigm and a new meanly-unsupervised method, that is more biologically inspired. It tries to gain more insights from the computational neuroscience community in order to incorporate concepts like time, context and attention during the learning process which are typical of the human brain. In the end, the thesis is supposed to prove that in certain cases, with a lower quantity of data, HTM can outperform CNN.

Named Entity Extraction: la piattaforma Gate/Annie

In questo lavoro si introducono i concetti di base di Natural Language Processing, soffermandosi su Information Extraction e analizzandone gli ambiti applicativi, le attività principali e la differenza rispetto a Information Retrieval. Successivamente si analizza il processo di Named Entity Recognition, focalizzando l’attenzione sulle principali problematiche di annotazione di testi e sui metodi per la valutazione della qualità dell’estrazione di entità. Infine si fornisce una panoramica della piattaforma software open-source di language processing GATE/ANNIE, descrivendone l’architettura e i suoi componenti principali, con approfondimenti sugli strumenti che GATE offre per l'approccio rule-based a Named Entity Recognition.

Un’applicazione di elaborazione del linguaggio naturale per il gioco di "Indovina chi?" su robot umanoide

La tesi è stata incentrata sul gioco «Indovina chi?» per l’identificazione da parte del robot Nao di un personaggio tramite la sua descrizione. In particolare la descrizione avviene tramite domande e risposte L’obiettivo della tesi è la progettazione di un sistema in grado di capire ed elaborare dei dati comunicati usando un sottoinsieme del linguaggio naturale, estrapolarne le informazioni chiave e ottenere un riscontro con informazioni date in precedenza. Si è quindi programmato il robot Nao in modo che sia in grado di giocare una partita di «Indovina chi?» contro un umano comunicando tramite il linguaggio naturale. Sono state implementate regole di estrazione e categorizzazione per la comprensione del testo utilizzando Cogito, una tecnologia brevettata dall'azienda Expert System. In questo modo il robot è in grado di capire le risposte e rispondere alle domande formulate dall'umano mediante il linguaggio naturale. Per il riconoscimento vocale è stata utilizzata l'API di Google e PyAudio per l'utilizzo del microfono. Il programma è stato implementato in Python e i dati dei personaggi sono memorizzati in un database che viene interrogato e modificato dal robot. L'algoritmo del gioco si basa su calcoli probabilistici di vittoria del robot e sulla scelta delle domande da proporre in base alle risposte precedentemente ricevute dall'umano. Le regole semantiche realizzate danno la possibilità al giocatore di formulare frasi utilizzando il linguaggio naturale, inoltre il robot è in grado di distinguere le informazioni che riguardano il personaggio da indovinare senza farsi ingannare. La percentuale di vittoria del robot ottenuta giocando 20 partite è stata del 50%. Il data base è stato sviluppato in modo da poter realizzare un identikit completo di una persona, oltre a quello dei personaggi del gioco. È quindi possibile ampliare il progetto per altri scopi, oltre a quello del gioco, nel campo dell'identificazione.

Towards Automatic Identification of Discourse Markers in Dialogs: the Case of 'Like'

This article discusses the detection of discourse markers (DM) in dialog transcriptions, by human annotators and by automated means. After a theoretical discussion of the definition of DMs and their relevance to natural language processing, we focus on the role of like as a DM. Results from experiments with human annotators show that detection of DMs is a difficult but reliable task, which requires prosodic information from soundtracks. Then, several types of features are defined for automatic disambiguation of like: collocations, part-of-speech tags and duration-based features. Decision-tree learning shows that for like, nearly 70% precision can be reached, with near 100% recall, mainly using collocation filters. Similar results hold for well, with about 91% precision at 100% recall.

La recuperación de información en el siglo XX : Revisión y aplicación de aspectos de la lingüística cuantitativa y la modelización matemática de la información

Esta tesina indaga en el ámbito de las Tecnologías de la Información sobre los diferentes desarrollos realizados en la interpretación automática de la semántica de textos y su relación con los Sistemas de Recuperación de Información. Partiendo de una revisión bibliográfica selectiva se busca sistematizar la documentación estableciendo de manera evolutiva los principales antecedentes y técnicas, sintetizando los conceptos fundamentales y resaltando los aspectos que justifican la elección de unos u otros procedimientos en la resolución de los problemas.

La recuperación de información en el siglo XX : Revisión y aplicación de aspectos de la lingüística cuantitativa y la modelización matemática de la información

Esta tesina indaga en el ámbito de las Tecnologías de la Información sobre los diferentes desarrollos realizados en la interpretación automática de la semántica de textos y su relación con los Sistemas de Recuperación de Información. Partiendo de una revisión bibliográfica selectiva se busca sistematizar la documentación estableciendo de manera evolutiva los principales antecedentes y técnicas, sintetizando los conceptos fundamentales y resaltando los aspectos que justifican la elección de unos u otros procedimientos en la resolución de los problemas.

La recuperación de información en el siglo XX : Revisión y aplicación de aspectos de la lingüística cuantitativa y la modelización matemática de la información

Esta tesina indaga en el ámbito de las Tecnologías de la Información sobre los diferentes desarrollos realizados en la interpretación automática de la semántica de textos y su relación con los Sistemas de Recuperación de Información. Partiendo de una revisión bibliográfica selectiva se busca sistematizar la documentación estableciendo de manera evolutiva los principales antecedentes y técnicas, sintetizando los conceptos fundamentales y resaltando los aspectos que justifican la elección de unos u otros procedimientos en la resolución de los problemas.

Essentials in Ontology Engineering: Methodologies, Languages, and Tools

In the beginning of the 90s, ontology development was similar to an art: ontology developers did not have clear guidelines on how to build ontologies but only some design criteria to be followed. Work on principles, methods and methodologies, together with supporting technologies and languages, made ontology development become an engineering discipline, the so-called Ontology Engineering. Ontology Engineering refers to the set of activities that concern the ontology development process and the ontology life cycle, the methods and methodologies for building ontologies, and the tool suites and languages that support them. Thanks to the work done in the Ontology Engineering field, the development of ontologies within and between teams has increased and improved, as well as the possibility of reusing ontologies in other developments and in final applications. Currently, ontologies are widely used in (a) Knowledge Engineering, Artificial Intelligence and Computer Science, (b) applications related to knowledge management, natural language processing, e-commerce, intelligent information integration, information retrieval, database design and integration, bio-informatics, education, and (c) the Semantic Web, the Semantic Grid, and the Linked Data initiative. In this paper, we provide an overview of Ontology Engineering, mentioning the most outstanding and used methodologies, languages, and tools for building ontologies. In addition, we include some words on how all these elements can be used in the Linked Data initiative.

Diseño e implementación de un algoritmo para la detección de la negación de textos clínicos en español

Vivimos en una época en la que cada vez existe una mayor cantidad de información. En el dominio de la salud la historia clínica digital ha permitido digitalizar toda la información de los pacientes. Estas historias clínicas digitales contienen una gran cantidad de información valiosa escrita en forma narrativa que sólo podremos extraer recurriendo a técnicas de procesado de lenguaje natural. No obstante, si se quiere realizar búsquedas sobre estos textos es importante analizar que la información relativa a síntomas, enfermedades, tratamientos etc. se puede refererir al propio paciente o a sus antecentes familiares, y que ciertos términos pueden aparecer negados o ser hipotéticos. A pesar de que el español ocupa la segunda posición en el listado de idiomas más hablados con más de 500 millones de hispano hablantes, hasta donde tenemos de detección de la negación, probabilidad e histórico en textos clínicos en español. Por tanto, este Trabajo Fin de Grado presenta una implementación basada en el algoritmo ConText para la detección de la negación, probabilidad e histórico en textos clínicos escritos en español. El algoritmo se ha validado con 454 oraciones que incluían un total de 1897 disparadores obteniendo unos resultado de 83.5 %, 96.1 %, 96.9 %, 99.7% y 93.4% de exactitud con condiciones afirmados, negados, probable, probable negado e histórico respectivamente. ---ABSTRACT---We live in an era in which there is a huge amount of information. In the domain of health, the electronic health record has allowed to digitize all the information of the patients. These electronic health records contain valuable information written in narrative form that can only be extracted using techniques of natural language processing. However, if you want to search on these texts is important to analyze if the relative information about symptoms, diseases, treatments, etc. are referred to the patient or family casework, and that certain terms may appear negated or be hypothesis. Although Spanish is the second spoken language with more than 500 million speakers, there seems to be no method of detection of negation, hypothesis or historical in medical texts written in Spanish. Thus, this bachelor’s final degree presents an implementation based on the ConText algorithm for the detection of negation, hypothesis and historical in medical texts written in Spanish. The algorithm has been validated with 454 sentences that included a total of 1897 triggers getting a result of 83.5 %, 96.1 %, 96.9 %, 99.7% and 93.4% accuracy with affirmed, negated, hypothesis, negated hypothesis and historical respectively.

Sistema de recomendación semántico para la movilidad en el entorno AFRICA BUILD Portal

La Gestión de Recursos Humanos a través de Internet es un problema latente y presente actualmente en cualquier sitio web dedicado a la búsqueda de empleo. Este problema también está presente en AFRICA BUILD Portal. AFRICA BUILD Portal es una emergente red socio-profesional nacida con el ánimo de crear comunidades virtuales que fomenten la educación e investigación en el área de la salud en países africanos. Uno de los métodos para fomentar la educación e investigación es mediante la movilidad de estudiantes e investigadores entre instituciones, apareciendo así, el citado problema de la gestión de recursos humanos. Por tanto, este trabajo se centra en solventar el problema de la gestión de recursos humanos en el entorno específico de AFRICA BUILD Portal. Para solventar este problema, el objetivo es desarrollar un sistema de recomendación que ayude en la gestión de recursos humanos en lo que concierne a la selección de las mejores ofertas y demandas de movilidad. Caracterizando al sistema de recomendación como un sistema semántico el cual ofrecerá las recomendaciones basándose en las reglas y restricciones impuestas por el dominio. La aproximación propuesta se basa en seguir el enfoque de los sistemas de Matchmaking semánticos. Siguiendo este enfoque, por un lado, se ha empleado un razonador de lógica descriptiva que ofrece inferencias útiles en el cálculo de las recomendaciones y por otro lado, herramientas de procesamiento de lenguaje natural para dar soporte al proceso de recomendación. Finalmente para la integración del sistema de recomendación con AFRICA BUILD Portal se han empleado diversas tecnologías web. Los resultados del sistema basados en la comparación de recomendaciones creadas por el sistema y por usuarios reales han mostrado un funcionamiento y rendimiento aceptable. Empleando medidas de evaluación de sistemas de recuperación de información se ha obtenido una precisión media del sistema de un 52%, cifra satisfactoria tratándose de un sistema semántico. Pudiendo concluir que con la solución implementada se ha construido un sistema estable y modular posibilitando: por un lado, una fácil evolución que debería ir encaminada a lograr un rendimiento mayor, incrementando su precisión y por otro lado, dejando abiertas nuevas vías de crecimiento orientadas a la explotación del potencial de AFRICA BUILD Portal mediante la Web 3.0. ---ABSTRACT---The Human Resource Management through Internet is currently a latent problem shown in any employment website. This problem has also appeared in AFRICA BUILD Portal. AFRICA BUILD Portal is an emerging socio-professional network with the objective of creating virtual communities to foster the capacity for health research and education in African countries. One way to foster this capacity of research and education is through the mobility of students and researches between institutions, thus appearing the Human Resource Management problem. Therefore, this dissertation focuses on solving the Human Resource Management problem in the specific environment of AFRICA BUILD Portal. To solve this problem, the objective is to develop a recommender system which assists the management of Human Resources with respect to the selection of the best mobility supplies and demands. The recommender system is a semantic system which will provide the recommendations according to the domain rules and restrictions. The proposed approach is based on semantic matchmaking solutions. So, this approach on the one hand uses a Description Logics reasoning engine which provides useful inferences to the recommendation process and on the other hand uses Natural Language Processing techniques to support the recommendation process. Finally, Web technologies are used in order to integrate the recommendation system into AFRICA BUILD Portal. The results of evaluating the system are based on the comparison between recommendations created by the system and by real users. These results have shown an acceptable behavior and performance. The average precision of the system has been obtained by evaluation measures for information retrieval systems, so the average precision of the system is at 52% which may be considered as a satisfactory result taking into account that the system is a semantic system. To conclude, it could be stated that the implemented system is stable and modular. This fact on the one hand allows an easy evolution that should aim to achieve a higher performance by increasing its average precision and on the other hand keeps open new ways to increase the functionality of the system oriented to exploit the potential of AFRICA BUILD Portal through Web 3.0.