Técnicas de machine learning para la detección de la negación en textos clínicos en español

En los últimos años han surgido nuevos campos de las tecnologías de la información que exploran el tratamiento de la gran cantidad de datos digitales existentes y cómo transformarlos en conocimiento explícito. Las técnicas de Procesamiento del Lenguaje Natural (NLP) son capaces de extraer información de los textos digitales presentados en forma narrativa. Además, las técnicas de machine learning clasifican instancias o ejemplos en función de sus atributos, en distintas categorías, aprendiendo de otros previamente clasificados. Los textos clínicos son una gran fuente de información no estructurada; en consecuencia, información no explotada en su totalidad. Algunos términos usados en textos clínicos se encuentran en una situación de afirmación, negación, hipótesis o histórica. La detección de esta situación es necesaria para la estructuración de información, pero a su vez tiene una gran complejidad. Extrayendo características lingüísticas de los elementos, o tokens, de los textos mediante NLP; transformando estos tokens en instancias y las características en atributos, podemos mediante técnicas de machine learning clasificarlos con el objetivo de detectar si se encuentran afirmados, negados, hipotéticos o históricos. La selección de los atributos que cada token debe tener para su clasificación, así como la selección del algoritmo de machine learning utilizado son elementos cruciales para la clasificación. Son, de hecho, los elementos que componen el modelo de clasificación. Consecuentemente, este trabajo aborda el proceso de extracción de características, selección de atributos y selección del algoritmo de machine learning para la detección de la negación en textos clínicos en español. Se expone un modelo para la clasificación que, mediante el algoritmo J48 y 35 atributos obtenidos de características lingüísticas (morfológicas y sintácticas) y disparadores de negación, detecta si un token está negado en 465 frases provenientes de textos clínicos con un F-Score del 73%, una exhaustividad del 66% y una precisión del 81% con una validación cruzada de 10 iteraciones. ---ABSTRACT--- New information technologies have emerged in the recent years which explore the processing of the huge amount of existing digital data and its transformation into knowledge. Natural Language Processing (NLP) techniques are able to extract certain features from digital texts. Additionally, through machine learning techniques it is feasible to classify instances according to different categories, learning from others previously classified. Clinical texts contain great amount of unstructured data, therefore information not fully exploited. Some terms (tokens) in clinical texts appear in different situations such as affirmed, negated, hypothetic or historic. Detecting this situation is necessary for the structuring of this data, however not simple. It is possible to detect whether if a token is negated, affirmed, hypothetic or historic by extracting its linguistic features by NLP; transforming these tokens into instances, the features into attributes, and classifying these instances through machine learning techniques. Selecting the attributes each instance must have, and choosing the machine learning algorithm are crucial issues for the classification. In fact, these elements set the classification model. Consequently, this work approaches the features retrieval as well as the attributes and algorithm selection process used by machine learning techniques for the detection of negation in clinical texts in Spanish. We present a classification model which, through J48 algorithm and 35 attributes from linguistic features (morphologic and syntactic) and negation triggers, detects whether if a token is negated in 465 sentences from historical records, with a result of 73% FScore, 66% recall and 81% precision using a 10-fold cross-validation.

VHDL-based system design of a cognitive sensorimotor loop (CSL) for haptic Human-Machine Interaction (HMI)

This document is a summary of the Bachelor thesis titled “VHDL-Based System Design of a Cognitive Sensorimotor Loop (CSL) for Haptic Human-Machine Interaction (HMI)” written by Pablo de Miguel Morales, Electronics Engineering student at the Universidad Politécnica de Madrid (UPM Madrid, Spain) during an Erasmus+ Exchange Program at the Beuth Hochschule für Technik (BHT Berlin, Germany). The tutor of this project is Dr. Prof. Hild. This project has been developed inside the Neurobotics Research Laboratory (NRL) in close collaboration with Benjamin Panreck, a member of the NRL, and another exchange student from the UPM Pablo Gabriel Lezcano. For a deeper comprehension of the content of the thesis, a deeper look in the document is needed as well as the viewing of the videos and the VHDL design. In the growing field of automation, a large amount of workforce is dedicated to improve, adapt and design motor controllers for a wide variety of applications. In the specific field of robotics or other machinery designed to interact with humans or their environment, new needs and technological solutions are often being discovered due to the existing, relatively unexplored new scenario it is. The project consisted of three main parts: Two VHDL-based systems and one short experiment on the haptic perception. Both VHDL systems are based on a Cognitive Sensorimotor Loop (CSL) which is a control loop designed by the NRL and mainly developed by Dr. Prof. Hild. The CSL is a control loop whose main characteristic is the fact that it does not use any external sensor to measure the speed or position of the motor but the motor itself. The motor always generates a voltage that is proportional to its angular speed so it does not need calibration. This method is energy efficient and simplifies control loops in complex systems. The first system, named CSL Stay In Touch (SIT), consists in a one DC motor system controller by a FPGA Board (Zynq ZYBO 7000) whose aim is to keep contact with any external object that touches its Sensing Platform in both directions. Apart from the main behavior, three features (Search Mode, Inertia Mode and Return Mode) have been designed to enhance the haptic interaction experience. Additionally, a VGA-Screen is also controlled by the FPGA Board for the monitoring of the whole system. This system has been completely developed, tested and improved; analyzing its timing and consumption properties. The second system, named CSL Fingerlike Mechanism (FM), consists in a fingerlike mechanical system controlled by two DC motors (Each controlling one part of the finger). The behavior is similar to the first system but in a more complex structure. This system was optional and not part of the original objectives of the thesis and it could not be properly finished and tested due to the lack of time. The haptic perception experiment was an experiment conducted to have an insight into the complexity of human haptic perception in order to implement this knowledge into technological applications. The experiment consisted in testing the capability of the subjects to recognize different objects and shapes while being blindfolded and with their ears covered. Two groups were done, one had full haptic perception while the other had to explore the environment with a plastic piece attached to their finger to create a haptic handicap. The conclusion of the thesis was that a haptic system based only on a CSL-based system is not enough to retrieve valuable information from the environment and that other sensors are needed (temperature, pressure, etc.) but that a CSL-based system is very useful to control the force applied by the system to interact with haptic sensible surfaces such as skin or tactile screens. RESUMEN. Este documento es un resumen del proyecto fin de grado titulado “VHDL-Based System Design of a Cognitive Sensorimotor Loop (CSL) for Haptic Human-Machine Interaction (HMI)” escrito por Pablo de Miguel, estudiante de Ingeniería Electrónica de Comunicaciones en la Universidad Politécnica de Madrid (UPM Madrid, España) durante un programa de intercambio Erasmus+ en la Beuth Hochschule für Technik (BHT Berlin, Alemania). El tutor de este proyecto ha sido Dr. Prof. Hild. Este proyecto se ha desarrollado dentro del Neurorobotics Research Laboratory (NRL) en estrecha colaboración con Benjamin Panreck (un miembro del NRL) y con Pablo Lezcano (Otro estudiante de intercambio de la UPM). Para una comprensión completa del trabajo es necesaria una lectura detenida de todo el documento y el visionado de los videos y análisis del diseño VHDL incluidos en el CD adjunto. En el creciente sector de la automatización, una gran cantidad de esfuerzo está dedicada a mejorar, adaptar y diseñar controladores de motor para un gran rango de aplicaciones. En el campo específico de la robótica u otra maquinaria diseñada para interactuar con los humanos o con su entorno, nuevas necesidades y soluciones tecnológicas se siguen desarrollado debido al relativamente inexplorado y nuevo escenario que supone. El proyecto consta de tres partes principales: Dos sistemas basados en VHDL y un pequeño experimento sobre la percepción háptica. Ambos sistemas VHDL están basados en el Cognitive Sesnorimotor Loop (CSL) que es un lazo de control creado por el NRL y cuyo desarrollador principal ha sido Dr. Prof. Hild. El CSL es un lazo de control cuya principal característica es la ausencia de sensores externos para medir la velocidad o la posición del motor, usando el propio motor como sensor. El motor siempre genera un voltaje proporcional a su velocidad angular de modo que no es necesaria calibración. Este método es eficiente en términos energéticos y simplifica los lazos de control en sistemas complejos. El primer sistema, llamado CSL Stay In Touch (SIT), consiste en un sistema formado por un motor DC controlado por una FPGA Board (Zynq ZYBO 7000) cuyo objetivo es mantener contacto con cualquier objeto externo que toque su plataforma sensible en ambas direcciones. Aparte del funcionamiento básico, tres modos (Search Mode, Inertia Mode y Return Mode) han sido diseñados para mejorar la interacción. Adicionalmente, se ha diseñado el control a través de la FPGA Board de una pantalla VGA para la monitorización de todo el sistema. El sistema ha sido totalmente desarrollado, testeado y mejorado; analizando su propiedades de timing y consumo energético. El segundo sistema, llamado CSL Fingerlike Mechanism (FM), consiste en un mecanismo similar a un dedo controlado por dos motores DC (Cada uno controlando una falange). Su comportamiento es similar al del primer sistema pero con una estructura más compleja. Este sistema no formaba parte de los objetivos iniciales del proyecto y por lo tanto era opcional. No pudo ser plenamente desarrollado debido a la falta de tiempo. El experimento de percepción háptica fue diseñado para profundizar en la percepción háptica humana con el objetivo de aplicar este conocimiento en aplicaciones tecnológicas. El experimento consistía en testear la capacidad de los sujetos para reconocer diferentes objetos, formas y texturas en condiciones de privación del sentido del oído y la vista. Se crearon dos grupos, en uno los sujetos tenían plena percepción háptica mientras que en el otro debían interactuar con los objetos a través de una pieza de plástico para generar un hándicap háptico. La conclusión del proyecto fue que un sistema háptico basado solo en sistemas CSL no es suficiente para recopilar información valiosa del entorno y que debe hacer uso de otros sensores (temperatura, presión, etc.). En cambio, un sistema basado en CSL es idóneo para el control de la fuerza aplicada por el sistema durante la interacción con superficies hápticas sensibles tales como la piel o pantallas táctiles.

Structure-activity relationships derived by machine learning: the use of atoms and their bond connectivities to predict mutagenicity by inductive logic programming.

We present a general approach to forming structure-activity relationships (SARs). This approach is based on representing chemical structure by atoms and their bond connectivities in combination with the inductive logic programming (ILP) algorithm PROGOL. Existing SAR methods describe chemical structure by using attributes which are general properties of an object. It is not possible to map chemical structure directly to attribute-based descriptions, as such descriptions have no internal organization. A more natural and general way to describe chemical structure is to use a relational description, where the internal construction of the description maps that of the object described. Our atom and bond connectivities representation is a relational description. ILP algorithms can form SARs with relational descriptions. We have tested the relational approach by investigating the SARs of 230 aromatic and heteroaromatic nitro compounds. These compounds had been split previously into two subsets, 188 compounds that were amenable to regression and 42 that were not. For the 188 compounds, a SAR was found that was as accurate as the best statistical or neural network-generated SARs. The PROGOL SAR has the advantages that it did not need the use of any indicator variables handcrafted by an expert, and the generated rules were easily comprehensible. For the 42 compounds, PROGOL formed a SAR that was significantly (P < 0.025) more accurate than linear regression, quadratic regression, and back-propagation. This SAR is based on an automatically generated structural alert for mutagenicity.

Scientific bases of human-machine communication by voice.

The scientific bases for human-machine communication by voice are in the fields of psychology, linguistics, acoustics, signal processing, computer science, and integrated circuit technology. The purpose of this paper is to highlight the basic scientific and technological issues in human-machine communication by voice and to point out areas of future research opportunity. The discussion is organized around the following major issues in implementing human-machine voice communication systems: (i) hardware/software implementation of the system, (ii) speech synthesis for voice output, (iii) speech recognition and understanding for voice input, and (iv) usability factors related to how humans interact with machines.

The role of voice input for human-machine communication.

Optimism is growing that the near future will witness rapid growth in human-computer interaction using voice. System prototypes have recently been built that demonstrate speaker-independent real-time speech recognition, and understanding of naturally spoken utterances with vocabularies of 1000 to 2000 words, and larger. Already, computer manufacturers are building speech recognition subsystems into their new product lines. However, before this technology can be broadly useful, a substantial knowledge base is needed about human spoken language and performance during computer-based spoken interaction. This paper reviews application areas in which spoken interaction can play a significant role, assesses potential benefits of spoken interaction with machines, and compares voice with other modalities of human-computer interaction. It also discusses information that will be needed to build a firm empirical foundation for the design of future spoken and multimodal interfaces. Finally, it argues for a more systematic and scientific approach to investigating spoken input and performance with future language technology.

Research in speech communication.

Advances in digital speech processing are now supporting application and deployment of a variety of speech technologies for human/machine communication. In fact, new businesses are rapidly forming about these technologies. But these capabilities are of little use unless society can afford them. Happily, explosive advances in microelectronics over the past two decades have assured affordable access to this sophistication as well as to the underlying computing technology. The research challenges in speech processing remain in the traditionally identified areas of recognition, synthesis, and coding. These three areas have typically been addressed individually, often with significant isolation among the efforts. But they are all facets of the same fundamental issue--how to represent and quantify the information in the speech signal. This implies deeper understanding of the physics of speech production, the constraints that the conventions of language impose, and the mechanism for information processing in the auditory system. In ongoing research, therefore, we seek more accurate models of speech generation, better computational formulations of language, and realistic perceptual guides for speech processing--along with ways to coalesce the fundamental issues of recognition, synthesis, and coding. Successful solution will yield the long-sought dictation machine, high-quality synthesis from text, and the ultimate in low bit-rate transmission of speech. It will also open the door to language-translating telephony, where the synthetic foreign translation can be in the voice of the originating talker.

Military and government applications of human-machine communication by voice.

This paper describes a range of opportunities for military and government applications of human-machine communication by voice, based on visits and contacts with numerous user organizations in the United States. The applications include some that appear to be feasible by careful integration of current state-of-the-art technology and others that will require a varying mix of advances in speech technology and in integration of the technology into applications environments. Applications that are described include (1) speech recognition and synthesis for mobile command and control; (2) speech processing for a portable multifunction soldier's computer; (3) speech- and language-based technology for naval combat team tactical training; (4) speech technology for command and control on a carrier flight deck; (5) control of auxiliary systems, and alert and warning generation, in fighter aircraft and helicopters; and (6) voice check-in, report entry, and communication for law enforcement agents or special forces. A phased approach for transfer of the technology into applications is advocated, where integration of applications systems is pursued in parallel with advanced research to meet future needs.

Deployment of human-machine dialogue systems.

The deployment of systems for human-to-machine communication by voice requires overcoming a variety of obstacles that affect the speech-processing technologies. Problems encountered in the field might include variation in speaking style, acoustic noise, ambiguity of language, or confusion on the part of the speaker. The diversity of these practical problems encountered in the "real world" leads to the perceived gap between laboratory and "real-world" performance. To answer the question "What applications can speech technology support today?" the concept of the "degree of difficulty" of an application is introduced. The degree of difficulty depends not only on the demands placed on the speech recognition and speech synthesis technologies but also on the expectations of the user of the system. Experience has shown that deployment of effective speech communication systems requires an iterative process. This paper discusses general deployment principles, which are illustrated by several examples of human-machine communication systems.

Calix[6]arene Derivatives as Pivotal Components of Working Devices and Molecular Machine Prototypes

This Ph.D. thesis describes the synthesis, characterization and study of calix[6]arene derivatives as pivotal components for the construction of molecular machine prototypes. Initially, the ability of a calix[6]arene wheel to supramolecularly assist and increase the rate of a nucleophilic substitution reaction was exploited for the synthesis of two constitutionally isomeric oriented rotaxanes. Then, the synthesis and characterization of several hetero-functionalised calix[6]arene derivatives and the possibility to obtain molecular muscle prototypes was reported. The ability of calix[6]arenes to form oriented pseudorotaxane towards dialkyl viologen axles was then exploited for the synthesis of two calixarene-based [2]catenanes. As last part of this thesis, studies on the electrochemical response of the threading-dethreading process of calix[6]arene-based pseudorotaxanes and rotaxanes supported on glassy carbon electrodes are reported.

Translating and Subtitling Humor Through Theory and Practice

This project assesses translating and subtitling humor in Italian and Spanish language films subtitled into English. Humor in film is problematic to translate when subtitling: visual humor may need no assistance to be delivered to a target audience, but verbal humor requires thorough analysis to be constructed effectively in the target language. To keep humor alive in target language translations, translators must understand the structure and function of humor. This project researches humor theory, translation and subtitling. It analyzes humor function through humor theory and applies this knowledge to translating audiovisual mediums. An understanding of joke structure and humor function can serve as a guide for translators to recognize, devise and evaluate equivalent translations of humor in film.

The Need for Professional Translation and Interpretation in School Districts

The language barrier prevents Latino students from experiencing academic success, and prevents Latino parents from participating in their children's education. Through a review of journal articles, research projects, doctoral dissertations, legislation, and books, this project studies the benefits and dangers of various methods of translating and interpreting in the education system, including issues created by language barriers in schools, common methods of translating and interpreting, and legislation addressing language barriers and education. The project reveals that schools use various methods to translate and interpret, including relying on children, school staff and machine translation, although such methods are often problematic and inaccurate. The project also reveals that professional translation and interpretation are superior to the various non-professional methods.

Machine Translation: Translators' Friend or Foe?

This paper discusses the impact of machine translation on the language industry, specifically addressing its effect on translators. It summarizes the history of the development of machine translation, explains the underlying theory that ties machine translation to its practical applications, and describes the different types of machine translation as well as other tools familiar to translators. There are arguments for and against its use, as well as evaluation methods for testing it. Internet and real-time communication are featured for their role in the increase of machine translation use. The potential that this technology has in the future of professional translation is examined. This paper shows that machine translation will continue to be increasingly used whether translators like it or not.

Machine learning techniques for automatic opinion detection in non-traditional textual genres

This paper presents a preliminary study in which Machine Learning experiments applied to Opinion Mining in blogs have been carried out. We created and annotated a blog corpus in Spanish using EmotiBlog. We evaluated the utility of the features labelled firstly carrying out experiments with combinations of them and secondly using the feature selection techniques, we also deal with several problems, such as the noisy character of the input texts, the small size of the training set, the granularity of the annotation scheme and the language object of our study, Spanish, with less resource than English. We obtained promising results considering that it is a preliminary study.