Atualmente, é comum que usuários levem em consideração a localização geográfica dos documentos, é dizer considerar o escopo geográfico que está sendo tratado no contexto do documento, nos processos de Recuperação de Informação. No entanto, os sistemas convencionais de extração de informação que estão baseados em palavras-chave não consideram que as palavras podem representar entidades geográficas espacialmente relacionadas com outras entidades nos documentos. Para resolver esse problema, é necessário viabilizar o georreferenciamento dos textos, ou seja, identificar as entidades geográficas presentes e associá-las com sua correta localização espacial. A identificação e desambiguação das entidades geográficas apresenta desafios importantes, principalmente do ponto de vista linguístico, já que um topônimo, pode possuir variados tipos de ambiguidade associados. Esse problema de ambiguidade causa ruido nos processos de recuperação de informação, já que o mesmo termo pode ter informação relevante ou irrelevante associada. Assim, a principal estratégia para superar os problemas de ambiguidade, compreende a identificação de evidências que auxiliem na identificação e desambiguação das localidades nos textos. O presente trabalho propõe uma metodologia que permite identificar e determinar a cobertura espacial dos documentos, denominada SpatialCIM. A metodologia SpatialCIM tem o objetivo de organizar os processos de resolução de topônimos. Assim, o principal objetivo deste trabalho é avaliar e selecionar técnicas de desambiguação que permitam resolver a ambiguidade dos topônimos nos textos. Para isso, foram propostas e desenvolvidas as abordagens de (1)Desambiguação por Pontos e a (2)Desambiguação Textual e Estrutural. Essas abordagens, exploram duas técnicas diferentes de desambiguação de topônimos, as quais, geram e desambiguam os caminhos geográficos associados aos topônimos reconhecidos para cada documento. Assim, a hipótese desta pesquisa é que o uso das técnicas de desambiguação de topônimos viabilizam uma melhor localização espacial dos documentos. A partir dos resultados obtidos neste trabalho, foi possível demonstrar que as técnicas de desambiguação melhoram a precisão e revocação na classificação espacial dos documentos. Demonstrou-se também o impacto positivo do uso de uma ferramenta linguística no processo de reconhecimento das entidades geográficas. Assim, foi demostrada a utilidade dos processos de desambiguação para a obtenção da cobertura espacial dos documentos / Currently, it is usual that users take into account the geographical localization of the documents in the Information Retrieval process. However, the conventional information retrieval systems based on key-word matching do not consider which words can represent geographical entities that are spatially related to other entities in the documents. To solve this problem, it is necessary to enable the geo-referencing of texts by identifying the geographical entities present in text and associate them with their correct spatial location. The identification and disambiguation of the geographical entities present major challenges mainly from the linguistic point of view, since one location can have different types of associated ambiguity. The ambiguity problem causes noise in the process of information retrieval, since the same term may have relevant or irrelevant information associated. Thus, the main strategy to overcome these problems, include the identification of evidence to assist in the identification and disambiguation of locations in the texts. This study proposes a methodology that allows the identification and spatial localization of the documents, denominated SpatialCIM. The SpatialCIM methodology has the objective to organize the Topônym Resolution process. Therefore the main objective of this study is to evaluate and select disambiguation techniques that allow solving the toponym ambiguity in texts. Therefore, we proposed and developed the approaches of (1) Disambiguation for Points and (2) Textual and Structural Disambiguation. These approaches exploit two different techniques of toponym disambiguation, which generate and desambiguate the associated paths with the recognized geographical toponym for each document. Therefore the hypothesis is, that the use of the toponyms disambiguation techniques enable a better spatial localization of documents. From the results it was possible to demonstrate that the disambiguation techniques improve the precision and recall for the spatial classification of documents. The positive effect of using a linguistic tool for the process of geographical entities recognition was also demonstrated. Thus, it was proved the usefulness of the disambiguation process for obtaining a spatial coverage of the document

Background: Traditionally text mention normalization corpora have normalized concepts to single ontology identifiers ("pre-coordinated concepts"). Less frequently, normalization corpora have used concepts with multiple identifiers ("post-coordinated concepts") but the additional identifiers have been restricted to a defined set of relationships to the core concept. This approach limits the ability of the normalization process to express semantic meaning. We generated a freely available corpus using post-coordinated concepts without a defined set of relationships that we term "compositional concepts" to evaluate their use in clinical text. Methods: We annotated 5397 disorder mentions from the ShARe corpus to SNOMED CT that were previously normalized as "CUI-less" in the "SemEval-2015 Task 14" shared task because they lacked a pre-coordinated mapping. Unlike the previous normalization method, we do not restrict concept mappings to a particular set of the Unified Medical Language System (UMLS) semantic types and allow normalization to occur to multiple UMLS Concept Unique Identifiers (CUIs). We computed annotator agreement and assessed semantic coverage with this method. Results: We generated the largest clinical text normalization corpus to date with mappings to multiple identifiers and made it freely available. All but 8 of the 5397 disorder mentions were normalized using this methodology. Annotator agreement ranged from 52.4% using the strictest metric (exact matching) to 78.2% using a hierarchical agreement that measures the overlap of shared ancestral nodes. Conclusion: Our results provide evidence that compositional concepts can increase semantic coverage in clinical text. To our knowledge we provide the first freely available corpus of compositional concept annotation in clinical text.

The goal of this master thesis is to design and implement a named entity recognition and linking algorithm. A part of this goal is to propose and create a knowledge base that will be used in the algorithm. Because of the limited amount of data for languages other than English, we want to be able to train our method on one language, and then transfer the learned parameters to other languages (that do not have enough training data). The thesis consists of description of available knowledge bases, existing methods and design and implementation of our own knowledge base and entity linking method. Our method achieves state of the art result on a few variants of the AIDA CoNLL-YAGO dataset. The method also obtains comparable results on a sample of Czech annotated data from the PDT dataset using the parameters trained on the English CoNLL dataset. Powered by TCPDF (www.tcpdf.org)

With the constantly growing amount of biomedical literature, methods for automatically distilling information from unstructured data, collectively known as information extraction, have become indispensable. Whilst most biomedical information extraction efforts in the last decade have focussed on the identification of gene products and interactions between them, the biomedical text mining community has recently extended their scope to capture associations between biomedical and chemical entities with the aim of supporting applications in drug discovery. This thesis is the first comprehensive study focussing on information extraction from pharmaceutical chemistry literature. In this research, we describe our work on (1) recognising names of chemical compounds and drugs, facilitated by the incorporation of domain knowledge; (2) exploring different coreference resolution paradigms in order to recognise co-referring expressions given a full-text article; and (3) defining drug-target interactions as events and distilling them from pharmaceutical chemistry literature using event extraction methods.

Cieľom tejto práce je preskúmať dostupné systémy pre extrakciu informácií a možnosti ich použitia v projekte MedIEQ. Teoretickú časť obsahuje úvod do oblasti extrakcie informácií. Popisujem účel, potreby a použitie a vzťah k iným úlohám spracovania prirodzeného jazyka. Prechádzam históriou, nedávnym vývojom, meraním výkonnosti a jeho kritikou. Taktiež popisujem všeobecnú architektúru IE systému a základné úlohy, ktoré má riešiť, s dôrazom na extrakciu entít. V praktickej časti sa nacházda prehľad algoritmov používaných v systémoch pre extrakciu informácií. Opisujem oba typy algoritmov ? pravidlové aj štatistické. V ďalšej kapitole je zoznam a krátky popis existujúcich voľných systémov. Nakoniec robím vlastný experiment s dvomi systémami ? LingPipe a GATE na vybraných korpusoch. Meriam rôzne výkonnostné štatistiky. Taktiež som vytvoril malý slovník a regulárny výraz pre email aby som demonštroval taktiež pravidlá pre extrahovanie určitých špecifických informácií.

Thesis Title: Named Entity Recognition in the Biomedical Domain Named entity recognition (NER) is the task of information extraction that attempts to recognize and extract particular entities in a text. One of the issues that stems from NER is that its models are domain specific. The goal of the thesis is to focus on entities strictly from the biomedical domain. The other issue with NER comes the synonymous terms that may be linked to one entity, moreover they lead to issue of disambiguation of the entities. Due to the popularity of neural networks and their success in NLP tasks, the work should use a neural network architecture for the task of named entity disambiguation, which is described in the paper by Eshel et al [1]. One of the subtasks of the thesis is to map the words and entities to a vector space using word embeddings, which attempts to provide textual context similarity, and coherence [2]. The main output of the thesis will be a model that attempts to disambiguate entities of the biomedical domain, using scientific journals (PubMed and Embase) as the documents of our interest.

An der Verarbeitung natürlicher Sprache durch computerbasierte Systeme wurde immer aktiv entwickelt und geforscht, um Aufgaben in den am weitesten verbreiteten Sprachen zu lösen. In dieser Arbeit werden verschiedene Ansätze zur Lösung von Problemen in diesem Bereich mittels künstlicher neuronaler Netze beschrieben. Dabei konzentriert sich diese Arbeit hauptsächlich auf modernere Architekturen wie Transformatoren oder BERT. Ziel dabei ist es, diese besser zu verstehen und herauszufinden, welche Vorteile sie gegenüber herkömmlichen künstlichen neuronalen Netzwerken haben. Anschließend wird dieses erlangte Wissen an einer Aufgabe aus dem Bereich der Verarbeitung natürlicher Sprache getestet, in welcher mittels einer sogenannten Named Entity Recognition (NER) spezielle Informationen aus Texten extrahiert werden.:1 Einleitung 1.1 Verarbeitung natürlicher Sprache (NLP) 1.2 Neuronale Netze 1.2.1 Biologischer Hintergrund 1.3 Aufbau der Arbeit 2 Grundlagen 2.1 Künstliche neuronale Netze 2.1.1 Arten des Lernens 2.1.2 Aktivierungsfunktionen 2.1.3 Verlustfunktionen 2.1.4 Optimierer 2.1.5 Über- und Unteranpassung 2.1.6 Explodierender und verschwindender Gradient 2.1.7 Optimierungsverfahren 3 Netzwerkarchitekturen zur Verarbeitung natürlicher Sprache 3.1 Rekurrente neuronale Netze (RNN) 3.1.1 Langes Kurzzeitgedächtnis (LSTM) 3.2 Autoencoder 3.3 Transformator 3.3.1 Worteinbettungen 3.3.2 Positionscodierung 3.3.3 Encoderblock 3.3.4 Decoderblock 3.3.5 Grenzen Transformatorarchitektur 3.4 Bidirektionale Encoder-Darstellungen von Transformatoren (BERT) 3.4.1 Vortraining 3.4.2 Feinabstimmung 4 Praktischer Teil und Ergebnisse 4.1 Aufgabe 4.2 Verwendete Bibliotheken, Programmiersprachen und Software 4.2.1 Python 4.2.2 NumPy 4.2.3 pandas 4.2.4 scikit-learn 4.2.5 Tensorflow 4.2.6 Keras 4.2.7 ktrain 4.2.8 Data Version Control (dvc) 4.2.9 FastAPI 4.2.10 Docker 4.2.11 Amazon Web Services 4.3 Daten 4.4 Netzwerkarchitektur 4.5 Training 4.6 Auswertung 4.7 Implementierung 5 Schlussbemerkungen 5.1 Zusammenfassung und Ausblick

The aim of this work is to find out which algorithm is the best at recognizing named entities in e-mail messages. The theoretical part explains the existing tools in this field. The practical part describes the design of two tools specifically designed to create new models capable of recognizing named entities in e-mail messages. The first tool is based on a neural network and the second tool uses a CRF graph model. The existing and newly created tools and their ability to generalize are compared on a subset of e-mail messages provided by Kiwi.com.

Digital gespeicherte Daten erfreuen sich einer stetig steigenden Verwendung. Insbesondere die computerbasierte Kommunikation über E-Mail, SMS, Messenger usw. hat klassische Kommunikationsmittel nahezu vollständig verdrängt. Einen Mehrwert aus diesen Daten zu generieren, ist sowohl im geschäftlichen als auch im privaten Bereich von entscheidender Bedeutung. Eine Möglichkeit den Nutzer zu unterstützen ist es, seine textuellen Daten umfassend zu analysieren und bestimmte Elemente hervorzuheben und ihm die Erstellung von Einträgen für Kalender, Adressbuch und dergleichen abzunehmen bzw. zumindest vorzubereiten. Eine weitere Möglichkeit stellt die semantische Suche in den Daten des Nutzers dar. Selbst mit Volltextsuche muss man bisher den genauen Wortlaut kennen, wenn man eine bestimmte Information sucht. Durch ein tiefgreifendes Verständnis für Zeit ist es nun aber möglich, über einen Zeitstrahl alle mit einem bestimmten Zeitpunkt oder einer Zeitspanne verknüpften Daten zu finden. Es existieren bereits viele Ansätze um Named Entity Recognition voll- bzw. semi-automatisch durchzuführen, aber insbesondere Verfahren, welche weitgehend sprachunabhängig arbeiten und sich somit leicht auf viele Sprachen skalieren lassen, sind kaum publiziert. Um ein solches Verfahren für natürlichsprachige Zeitausdrücke zu verbessern, werden in dieser Arbeit, basierend auf umfangreichen Analysen, Möglichkeiten vorgestellt. Es wird speziell eine Strategie entwickelt, die auf einem Verfahren des maschinellen Lernens beruht und so den manuellen Aufwand für die Unterstützung neuer Sprachen reduziert. Diese und weitere Strategien wurden implementiert und in die bestehende Architektur der Zeiterkennungsmaschine der ExB-Gruppe integriert.

This work researches named entity recognition (NER) with respect to images of documents with a domain-specific layout, by means of Convolutional Neural Networks (CNNs). Examples of such documents are receipts, invoices, forms and scientific papers, the latter of which are used in this work. An NER task is first performed statically, where a static number of entity classes is extracted per document. Networks based on the deep VGG-16 network are used for this task. Here, experimental evaluation shows that framing the task as a classification task, where the network classifies each bounding box coordinate separately, leads to the best network performance. Also, a multi-headed architecture is introduced, where the network has an independent fully-connected classification head per entity. VGG-16 achieves better performance with the multi-headed architecture than with its default, single-headed architecture. Additionally, it is shown that transfer learning does not improve performance of these networks. Analysis suggests that the networks trained for the static NER task learn to recognise document templates, rather than the entities themselves, and therefore do not generalize well to new, unseen templates. For a dynamic NER task, where the type and number of entity classes vary per document, experimental evaluation shows that, on large entities in the document, the Faster R-CNN object detection framework achieves comparable performance to the networks trained on the static task. Analysis suggests that Faster R-CNN generalizes better to new templates than the networks trained for the static task, as Faster R-CNN is trained on local features rather than the full document template. Finally, analysis shows that Faster R-CNN performs poorly on small entities in the image and suggestions are made to improve its performance.