Concept Annotation Research Articles

Visual pattern mining in histology image collections using bag of features

The paper addresses the problem of finding visual patterns in histology image collections. In particular, it proposes a method for correlating basic visual patterns with high-level concepts combining an appropriate image collection representation with state-of-the-art machine learning techniques. The proposed method starts by representing the visual content of the collection using a bag-of-features strategy. Then, two main visual mining tasks are performed: finding associations between visual-patterns and high-level concepts, and performing automatic image annotation. Associations are found using minimum-redundancy-maximum-relevance feature selection and co-clustering analysis. Annotation is done by applying a support-vector-machine classifier. Additionally, the proposed method includes an interpretation mechanism that associates concept annotations with corresponding image regions. The method was evaluated in two data sets: one comprising histology images from the different four fundamental tissues, and the other composed of histopathology images used for cancer diagnosis. Different visual-word representations and codebook sizes were tested. The performance in both concept association and image annotation tasks was qualitatively and quantitatively evaluated. The results show that the method is able to find highly discriminative visual features and to associate them to high-level concepts. In the annotation task the method showed a competitive performance: an increase of 21% in f-measure with respect to the baseline in the histopathology data set, and an increase of 47% in the histology data set. The experimental evidence suggests that the bag-of-features representation is a good alternative to represent visual content in histology images. The proposed method exploits this representation to perform visual pattern mining from a wider perspective where the focus is the image collection as a whole, rather than individual images.

Web Service Composition Algorithm

Web service composition is very important in web service technology. Domain Ontology is used for semantic web service match and automatic web service composition. When a web service is registered, it will be related to the concepts in the domain ontology with its output concepts. To start a semantic match or composition starts, only the related web service with the annotated concept will be used. The approach can deal with not only the sequential but also the parallel service composition problems The proposed approach was verified by experiments and case studies.

Sketch Based Video Annotation and Organization System in Distributed Teaching Environment

As the use of instructional video is becoming a key component of e-learning, there is an increasing need for a distributed system which supports collaborative video annotation and organization. In this paper, the authors construct a distributed environment on the top of NaradaBrokering to support collaborative operations on video material when users are located in different places. The concept of video annotation is enriched, making it a powerful media to improve the instructional video organizing and viewing. With panorama based and interpolation based methods, all related users can annotate or organize videos simultaneously. With these annotations, a video organization structure is consequently built through linking them with other video clips or annotations. Finally, an informal user study was conducted and result shows that this system improves the efficiency of video organizing and viewing and enhances user’s participating into the design process with good user experience.

Web image concept annotation with better understanding of tags and visual features

This paper focuses on improving the semi-manual method for web image concept annotation. By sufficiently studying the characteristics of tag and visual feature, we propose the Grouping-Based-Precision & Recall-Aided (GBPRA) feature selection strategy for concept annotation. Specifically, for visual features, we construct a more robust middle level feature by concatenating the k-NN results for each type of visual feature. For tag, we construct a concept-tag co-occurrence matrix, based on which the probability of an image belonging to certain concept can be calculated. By understanding the tags’ quality and groupings’ semantic depth, we propose a grouping based feature selection method; by studying the tags’ distribution, we adopt Precision and Recall as a complementary indicator for feature selection. In this way, the advantages of both tags and visual features are boosted. Experimental results show our method can achieve very high Average Precision, which greatly facilitates the annotation of large-scale web image dataset.

Annotation Based Query Answer over Inconsistent Database

In this paper, we introduce a concept of Annotation Based Query Answer, and a method for its computation, which can answer queries on relational databases that may violate a set of functional dependencies. In this approach, inconsistency is viewed as a property of data and described with annotations. To be more precise, every piece of data in a relation can have zero or more annotations with it and annotations are propagated along with queries from the source to the output. With annotations, inconsistent data in both input tables and query answers can be marked out but preserved, instead of being filtered in most previous work. Thus this approach can avoid information loss, a vital and common deficiency of most previous work in this area. To calculate query answers on an annotated database, we propose an algorithm to annotate the input tables, and redefine the five basic relational algebra operations (selection, projection, join, union and difference) so that annotations can be correctly propagated as the valid set of functional dependency changes during query processing. We also prove the soundness and completeness of the whole annotation computing system. Finally, we implement a prototype of our system, and give some performance experiments, which demonstrate that our approach is reasonable in running time, and excellent in information preserving.

Reconstruction annotation jamborees: a community approach to systems biology

Mol Syst Biol. 6: 361 Genome‐scale metabolic network reconstructions represent biochemical, genetic, and genomic (BiGG) knowledge bases for a target organism (Reed et al , 2006). Thus, they correspond to two‐dimensional genome annotations: that is, they contain all nodes and links that comprise a biochemical reaction network defined by the genome (Palsson, 2004). These reconstructions allow the conversion of biological knowledge into a mathematical format and subsequent computation of physiological properties. They therefore enable the formulation of a mechanistic genotype–phenotype relationship for metabolic functions in the target organism. The metabolic network reconstruction process is now well established (Thiele and Palsson, 2010) and its workflows have recently been reviewed (Reed et al , 2006; Feist et al , 2009). The development of a consensus network reconstruction that is accepted and used by the research community necessitates a collective effort to formalize such networks that are specific to a target organism. This need has led to the concept of a 2D annotation (or a reconstruction) jamboree (Mo and Palsson, 2009), in analogy to the 1D genome annotation jamborees that lead to a community‐driven genome annotation process. You may be interested in organizing a jamboree for your favorite target organism. What do you need to do? The goal of a network reconstruction jamboree is to reconcile and refine currently available BiGG knowledge about the target organism. If available, multiple existing metabolic network reconstructions made by individual research groups provide a great starting point. A jamboree should update, re‐evaluate, refine and, later on, expand the network content. These goals are most efficiently achieved through a community approach that assembles experts from different areas. Jamborees assist in fostering collaborations as well as informing the community about the properties, content, and capabilities of the consensus reconstruction to ensure its broad use for biological, biomedical, …

Conceptual language models for domain-specific retrieval

Over the years, various meta-languages have been used to manually enrich documents with conceptual knowledge of some kind. Examples include keyword assignment to citations or, more recently, tags to websites. In this paper we propose generative concept models as an extension to query modeling within the language modeling framework, which leverages these conceptual annotations to improve retrieval. By means of relevance feedback the original query is translated into a conceptual representation, which is subsequently used to update the query model. Extensive experimental work on five test collections in two domains shows that our approach gives significant improvements in terms of recall, initial precision and mean average precision with respect to a baseline without relevance feedback. On one test collection, it is also able to outperform a text-based pseudo-relevance feedback approach based on relevance models. On the other test collections it performs similarly to relevance models. Overall, conceptual language models have the added advantage of offering query and browsing suggestions in the form of conceptual annotations. In addition, the internal structure of the meta-language can be exploited to add related terms. Our contributions are threefold. First, an extensive study is conducted on how to effectively translate a textual query into a conceptual representation. Second, we propose a method for updating a textual query model using the concepts in conceptual representation. Finally, we provide an extensive analysis of when and how this conceptual feedback improves retrieval.

Developing a manually annotated clinical document corpus to identify phenotypic information for inflammatory bowel disease

BackgroundNatural Language Processing (NLP) systems can be used for specific Information Extraction (IE) tasks such as extracting phenotypic data from the electronic medical record (EMR). These data are useful for translational research and are often found only in free text clinical notes. A key required step for IE is the manual annotation of clinical corpora and the creation of a reference standard for (1) training and validation tasks and (2) to focus and clarify NLP system requirements. These tasks are time consuming, expensive, and require considerable effort on the part of human reviewers.MethodsUsing a set of clinical documents from the VA EMR for a particular use case of interest we identify specific challenges and present several opportunities for annotation tasks. We demonstrate specific methods using an open source annotation tool, a customized annotation schema, and a corpus of clinical documents for patients known to have a diagnosis of Inflammatory Bowel Disease (IBD). We report clinician annotator agreement at the document, concept, and concept attribute level. We estimate concept yield in terms of annotated concepts within specific note sections and document types.ResultsAnnotator agreement at the document level for documents that contained concepts of interest for IBD using estimated Kappa statistic (95% CI) was very high at 0.87 (0.82, 0.93). At the concept level, F-measure ranged from 0.61 to 0.83. However, agreement varied greatly at the specific concept attribute level. For this particular use case (IBD), clinical documents producing the highest concept yield per document included GI clinic notes and primary care notes. Within the various types of notes, the highest concept yield was in sections representing patient assessment and history of presenting illness. Ancillary service documents and family history and plan note sections produced the lowest concept yield.ConclusionChallenges include defining and building appropriate annotation schemas, adequately training clinician annotators, and determining the appropriate level of information to be annotated. Opportunities include narrowing the focus of information extraction to use case specific note types and sections, especially in cases where NLP systems will be used to extract information from large repositories of electronic clinical note documents.

An annotation model to reduce ambiguity in design communication

This paper describes an annotation model of digital design representations which aims to provide the design actors with a way to communicate more clearly in asynchronous situations. Our objective is to develop a model that improves design communication through the elicitation of knowledge related to the context of design. We present a case study based on our observations in an industrial vehicle company where we observed and analysed annotation practices during design reviews. The speech act theory is used in order to propose a model based on the concept of locutionary and illocutionary annotation acts, and annotation categories are drawn from cognitive approaches of design. An annotation scenario is then developed on the basis of the case study as an illustration of our model.

The SOL Genomics Network Model: Making Community Annotation Work

AbstractThe concept of community annotation is a growing discipline for achieving participation of the research community in depositing up‐to‐date knowledge in biological databases.The Solanaceae Genomics Network ("SGN":http://sgn.cornell.edu/) is a clade‐oriented database (COD) focusing on plants of the nightshade family, including tomato, potato, pepper, eggplant, and tobacco, and is one of the bioinformatics nodes of the international tomato genome sequencing project. One of our major efforts is linking Solanaceae phenotype information with the underlying genes, and subsequently the genome. As part of this goal, SGN has introduced a database for locus names and descriptors, and a database for phenotypes of natural and induced variation. These two databases have web interfaces that allow cross references, associations with tomato gene models, and in‐house curated information of sequences, literature, ontologies, gene networks, and the Solanaceae biochemical pathways database ("SolCyc":http://solcyc.sgn.cornell.edu). All of our curator tools are open for online community annotation, through specially assigned “submitter” accounts. Currently the community database consists of 5,548 phenotyped accessions, and 5,739 curated loci, out of which more than 300 loci where contributed or annotated by 66 active submitters, creating a database that is truly community driven.This framework is easily adaptable for other projects working on other taxa (for example see "http://chlamybase.org":http://chlamybase.org), greatly expanding the application of this user‐friendly online annotation system. Community participation is fostered by an active outreach program that includes contacting potential submitters via emails, at meetings and conferences, and by promoting featured user submitted annotations on the SGN homepage. The source code and database schema for all SGN functionalities are freely available. Please contact SGN at "sgn‐feedback[at]sgn.cornell.edu":mailto:sgn-feedback@sgn.cornell.edu for more information.

Assessment of disease named entity recognition on a corpus of annotated sentences

BackgroundIn recent years, the recognition of semantic types from the biomedical scientific literature has been focused on named entities like protein and gene names (PGNs) and gene ontology terms (GO terms). Other semantic types like diseases have not received the same level of attention. Different solutions have been proposed to identify disease named entities in the scientific literature. While matching the terminology with language patterns suffers from low recall (e.g., Whatizit) other solutions make use of morpho-syntactic features to better cover the full scope of terminological variability (e.g., MetaMap). Currently, MetaMap that is provided from the National Library of Medicine (NLM) is the state of the art solution for the annotation of concepts from UMLS (Unified Medical Language System) in the literature. Nonetheless, its performance has not yet been assessed on an annotated corpus. In addition, little effort has been invested so far to generate an annotated dataset that links disease entities in text to disease entries in a database, thesaurus or ontology and that could serve as a gold standard to benchmark text mining solutions.ResultsAs part of our research work, we have taken a corpus that has been delivered in the past for the identification of associations of genes to diseases based on the UMLS Metathesaurus and we have reprocessed and re-annotated the corpus. We have gathered annotations for disease entities from two curators, analyzed their disagreement (0.51 in the kappa-statistic) and composed a single annotated corpus for public use. Thereafter, three solutions for disease named entity recognition including MetaMap have been applied to the corpus to automatically annotate it with UMLS Metathesaurus concepts. The resulting annotations have been benchmarked to compare their performance.ConclusionsThe annotated corpus is publicly available at and can serve as a benchmark to other systems. In addition, we found that dictionary look-up already provides competitive results indicating that the use of disease terminology is highly standardized throughout the terminologies and the literature. MetaMap generates precise results at the expense of insufficient recall while our statistical method obtains better recall at a lower precision rate. Even better results in terms of precision are achieved by combining at least two of the three methods leading, but this approach again lowers recall. Altogether, our analysis gives a better understanding of the complexity of disease annotations in the literature. MetaMap and the dictionary based approach are available through the Whatizit web service infrastructure (Rebholz-Schuhmann D, Arregui M, Gaudan S, Kirsch H, Jimeno A: Text processing through Web services: Calling Whatizit. Bioinformatics 2008, 24:296-298).

Editorial and Retrospective 2008

Editorial and Retrospective 2008

The use of evolutionary biology concepts for genome annotation

The past decade has seen the completion of numerous whole-genome sequencing projects, began with bacterial genomes and continued with eukaryotic species from different phyla: fungi, plants and animals. Besides, more biological information are produced and are shared thanks to information exchange systems, and more biological concepts, as well as more bioinformatics tools, are available. In this article, we will describe how the evolutionary biology concepts, as well as computer science, are useful for a better understanding of biology in general and genome annotation in particular. The genome annotation process consists of taking the raw DNA produced, for example, by the genome sequencing projects, adding the layers of analysis and interpretation necessary to extract its biological significance and placing it in the context of our understanding of biological processes. Genome annotation is a multistep process falling into two broad categories: structural and functional annotation.

Personalized and adaptive navigation based on multimodal annotation

Although much effort is spent in developing navigation systems for pedestrians, many users with special needs are mostly excluded due to a lack of appropriate geographical data such as landmarks, waypoints, or obstacles. Such data is necessary for computing suitable routes because the best route might not be the shortest or fastest one. In this paper, the concept of multimodal annotation of geographical data for adaptive and personalized navigation is described. Direct input by the user is combined with data derived from the observation of the user's LOM-Modality (Location, Orientation, Movement) to annotate geographical data. Based on this data and data åderived from other users of the same user group, suitable routes even in unknown territory can be calculated.

Integration of knowledge management and collaborative learning by technical supported communication processes

To date the development of technical systems to support learning either in schools and universities (CSCL-systems), or in companies (knowledge-management systems) have largely developed separately concurrent, with no or little mutual exchange of experience. The approach presented here seeks to combine the strengths of both lines of development in order that both benefits from each others advantages achieve advantages for each development. Detailed requirements engineering is presented regarding, three research fields: communication theory; learning processes; and a study in companies. The requirements collected led to a sophisticated concept of annotations where annotations serve as communicative contributions, and (segmented) material is used as context. The combination and distinction between communicative contributions, and the material itself is a suitable approach for integrating the advantages of CSCL-systems (communication support) and knowledge management systems (support of content storage). These requirements are combined in a software system, KOLUMBUS, presented with its key features. To gather experience with KOLUMBUS and to convey potential for further improvement, two case studies were conducted. The main results, advantages and further potentials derived from these two case studies are described in this paper. The paper concludes with design recommendations for similar technical systems, as well as organizational hints for CSCL-groups (in schools or universities as well as in companies).

Concept-Based Speech-to-Speech Translation Using Maximum Entropy Models for Statistical Natural Concept Generation

The IBM Multilingual Automatic Speech-To-Speech TranslatOR (MASTOR) system is a research prototype developed for the Defense Advanced Research Projects Agency (DARPA) Babylon/CAST speech-to-speech machine translation program. The system consists of cascaded components of large-vocabulary conversational spontaneous speech recognition, statistical machine translation, and concatenative text-to-speech synthesis. To achieve highly accurate and robust conversational spoken language translation, a unique concept-based speech-to-speech translation approach is proposed that performs the translation by first understanding the meaning of the automatically recognized text. A decision-tree based statistical natural language understanding algorithm extracts the semantic information from the input sentences, while a natural language generation (NLG) algorithm predicts the translated text via maximum-entropy-based statistical models. One critical component in our statistical NLG approach is natural concept generation (NCG). The goal of NCG is not only to generate the correct set of concepts in the target language, but also to produce them in an appropriate order. To improve maximum-entropy-based concept generation, a set of new approaches is proposed. One approach improves concept sequence generation in the target language via forward–backward modeling, which selects the hypothesis with the highest combined conditional probability based on both the forward and backward generation models. This paradigm allows the exploration of both the left and right context information in the source and target languages during concept generation. Another approach selects bilingual features that enable maximum-entropy-based model training on the preannotated parallel corpora. This feature is augmented with word-level information in order to achieve higher NCG accuracy while minimizing the total number of distinct concepts and, hence, greatly reducing the concept annotation and natural language understanding effort. These features are further expanded to multiple sets to enhance model robustness. Finally, a confidence threshold is introduced to alleviate data sparseness problems in our training corpora. Experiments show a dramatic concept generation error rate reduction of more than 40% in our speech translation corpus within limited domains. Significant improvements of both word error rate and BiLingual Evaluation Understudy (BLEU) score are also achieved in our experiments on speech-to-speech translation.

The Jac System: Minimizing the Differences between Concurrent and Sequential Java Code.

JAC is a Java extension for high-level concurrent programming, meant to hide the notions of threads and synchronization statements from the programmer. Putting into practice the concept of concurrency annotations suggested for Eiffel some time ago, one of JAC's main assets is its support for minimizing the differences between concurrent and sequential implementations of objects and computations. The paper focuses on this aspect of JAC, presenting examples of successful applications of the annotations and pointing out remaining weaknesses.

Discovery of Web Services in a Multi-Ontology and Federated Registry Environment

The potential of a large-scale growth of private and semi-private registries is creating the need for an infrastructure that can support discovery and publication over a group of autonomous registries. Recent versions of UDDI have made changes to accommodate interactions between distributed registries. In this paper, we discuss an ontology-based Web service discovery infrastructure (METEOR-S Web Service Discovery Infrastructure), to provide access to registries that are divided based on business domains and grouped into federations. In addition, we discuss how Web service discovery is carried out within a federation. We provide a novel discovery algorithm, which addresses semantic heterogeneity with respect to multiple ontologies from the same domain. We also show through preliminary results of our empirical evaluation that even when services are annotated with different ontologies, our algorithm is able to find good matches and eliminate false matches by considering the context and the coverage information of the annotated concepts.

Constructing a Bowling Information System with Video Content Analysis

In this paper, we present a design and implementation of a bowling information system. This system contains three types of bowling game information including the bowling video content information, the game-related information and the player information. The MPEG-7 Description Schemes are used to describe these types of information and the relationships among them. This information is obtained through an annotator to which manual conceptual feature annotation (for the player and game-related information) and automatic perceptual feature extraction (for the video content information) are integrated. Several interesting events in the video such as strikes and the important frames are determined by automatically analyzing the video content. With an interactive user interface, users' queries are transformed into XQuery path expression to retrieve needed information about the bowling games to learn the skills of bowling. In addition to the implementation of the system, we also perform experiments to show the effectiveness of the automatic video content information extraction.

Quantifying Structure–Function Uncertainty: A Graph Theoretical Exploration into the Origins and Limitations of Protein Annotation

Since the advent of investigations into structural genomics, research has focused on correctly identifying domain boundaries, as well as domain similarities and differences in the context of their evolutionary relationships. As the science of structural genomics ramps up adding more and more information into the databanks, questions about the accuracy and completeness of our classification and annotation systems appear on the forefront of this research. A central question of paramount importance is how structural similarity relates to functional similarity. Here, we begin to rigorously and quantitatively answer these questions by first exploring the consensus between the most common protein domain structure annotation databases CATH, SCOP and FSSP. Each of these databases explores the evolutionary relationships between protein domains using a combination of automatic and manual, structural and functional, continuous and discrete similarity measures. In order to examine the issue of consensus thoroughly, we build a generalized graph out of each of these databases and hierarchically cluster these graphs at interval thresholds. We then employ a distance measure to find regions of greatest overlap. Using this procedure we were able not only to enumerate the level of consensus between the different annotation systems, but also to define the graph-theoretical origins behind the annotation schema of class, family and superfamily by observing that the same thresholds that define the best consensus regions between FSSP, SCOP and CATH correspond to distinct, non-random phase-transitions in the structure comparison graph itself. To investigate the correspondence in divergence between structure and function further, we introduce a measure of functional entropy that calculates divergence in function space. First, we use this measure to calculate the general correlation between structural homology and functional proximity. We extend this analysis further by quantitatively calculating the average amount of functional information gained from our understanding of structural distance and the corollary inherent uncertainty that represents the theoretical limit of our ability to infer function from structural similarity. Finally we show how our measure of functional “entropy” translates into a more intuitive concept of functional annotation into similarity EC classes.