Biomedical Natural Language Processing -- Editorial page -- Title page -- LCC data -- Acknowledgments -- Table of contents -- List of figures -- 1. Introduction to natural language processing -- 1.1 Some definitions -- 1.1.1 Computational linguistics -- 1.1.2 Natural language processing -- 1.1.3 Text mining -- 1.1.4 Usage of these definitions in practice -- 1.2 Levels of document and linguistic structure and their relationship to natural language processin -- 1.2.1 Document structure -- 1.2.2 Sentences -- 1.2.3 Tokens -- 1.2.4 Stems and lemmata -- 1.2.5 Part of speech -- 1.2.6 Syntactic structure -- 1.2.7 Semantics -- 2. Historical background -- 2.1 Early work in the medical domain -- 2.2 The emergence of the biological domain -- 2.3 Clinical text mining -- 2.4 Types of users of biomedical NLP systems -- 2.5 Resources and tools -- US National Library of Medicine -- MEDLINE database -- Medical Subject Headings -- PubMed -- GENIA -- PubMed Central International -- 2.6 Legal and ethical issues -- 2.7 Is biomedical natural language processing effective? -- 3. Named entity recognition -- 3.1 Overview -- 3.2 The crucial role of named entity recognition in BioNLP tasks -- 3.3 Why gene names are the way they are -- 3.4 An example of a rule-based gene NER system: KeX/PROPER -- 3.5 An example of a statistical disease NER system -- 3.6 Evaluation -- 4. Relation extraction -- 4.1 Introduction -- 4.1.1 Protein-protein interactions as an information extraction target -- 4.2 Binarity of most biomedical information extraction systems -- 4.3 Beyond simple binary relations -- 4.4 Rule-based systems -- 4.4.1 Co-occurrence -- 4.4.2 Example rule-based systems -- 4.4.3 Machine learning systems -- 4.5 Relations in clinical narrative -- 4.5.1 MedLEE -- 4.6 SemRep -- 4.6.1 NegEX -- 4.7 Evaluation -- 5. Information retrieval/document classification -- 5.1 Background.

5.1.1 Growth in the biomedical literature -- 5.1.2 PubMed/MEDLINE -- 5.2 Issues -- 5.3 A knowledge-based system that disambiguates gene names -- 5.4 A phrase-based search engine, with term and concept expansion and probabilistic relevance rankin -- 5.5 Full text -- 5.6 Image and figure search -- 5.7 Captions -- 5.7.1 Evaluation -- 6. Concept normalization -- 6.1 Gene normalization -- 6.1.1 The BioCreative definition of the gene normalization task -- 6.2 Building a successful gene normalization system -- 6.2.1 Coordination and ranges -- 6.2.2 An example system -- 6.3 Normalization and extraction of clinically pertinent terms -- 6.3.1 MetaMap UMLS mapping tools -- 7. Ontologies and computational lexical semantics -- 7.1 Unified Medical Language System (UMLS) -- 7.1.1 The Gene Ontology -- 7.2 Recognizing ontology terms in text -- 7.3 NLP for ontology quality assurance -- 7.4 Mapping, alignment, and linking of ontologies -- 8. Summarization -- 8.1 Medical summarization systems -- 8.1.1 Overview of medical summarization systems -- 8.1.2 A representative medical summarization system: Centrifuser -- 8.2 Genomics summarization systems -- 8.2.1 Sentence selection for protein-protein interactions -- 8.2.2 EntrezGene SUMMARY field generation -- 9. Question-answering -- 9.1 Principles -- 9.1.1 Question analysis and formal representation -- 9.1.1.1 Clinical questions -- 9.1.2 Formal representation of questions -- 9.1.3 Domain model-based question representation -- 9.1.3.1 Genomics and translational research questions -- 9.1.4 Answer retrieval -- 9.1.5 Answer extraction and generation -- 9.1.5.1 Reference answer formats for clinical questions -- 9.1.5.2 Entity-extraction approaches to answer generation -- 9.2 Applications -- 9.2.1 Question analysis and query formulation -- 9.2.2 Knowledge Extraction -- 9.2.2.1 Population Extractor -- 9.2.2.2 Problem Extractor.

9.2.2.3 Intervention Extractor -- 9.2.2.4 Outcome Extractor -- 9.2.2.5 Clinical Task classification -- 9.2.2.6 Strength of Evidence classification -- 9.2.2.7 Document scoring and ranking -- 9.2.3 Question-Document frame matching (PICO score) -- 9.2.3.1 Answer generation -- 9.2.4 Semantic clustering -- Summary -- 10. Software engineering -- 10.1 Introduction -- 10.2 Principles -- 10.3 General software testing -- 10.3.1 Clean and dirty tests -- 10.3.2 Testing requires planning -- 10.3.3 Catalogues -- 10.3.4 How many tests are possible? -- 10.3.5 Equivalence classes -- 10.3.6 Boundary conditions -- 10.4 Code coverage -- 10.5 When your input is language -- 10.6 User interface evaluation -- 10.6.1 API interface usability -- 11. Corpus construction and annotation -- 11.1 Corpora in the two domains as driving forces of research -- 11.2 Who should build biomedical corpora? -- 11.3 The relationship between annotation of entities and annotation of linguistic structure -- 11.4 Commonly used biomedical corpora -- 11.4.1 GENIA -- 11.4.2 CRAFT -- 11.4.3 BioCreative gene mention corpora -- 11.4.4 AIMed -- 11.4.5 Word sense disambiguation -- 11.4.6 Clinical corpora -- 11.4.6.1 NLP Challenge -- 11.4.6.2 The MIMIC collection -- 11.5 Factors that contribute to the success of biomedical corpora -- References -- Index.