Search

Purpose Publishing research data for reuse has become good practice in recent years. However, not much is known on how researchers actually find said data. In this exploratory study, we observe the information-seeking behaviour of social scientists searching for research data to reveal impediments and identify opportunities for data search infrastructure.Methods We asked 12 participants to search for research data and observed them in their natural environment. The sessions were recorded. Afterwards, we conducted semi-structured interviews to get a thorough understanding of their way of searching. From the recordings, we extracted the interaction behaviour of the participants and analysed the spoken words both during the search task and the interview by creating affinity diagrams.Results We found that literature search is more closely intertwined with dataset search than previous literature suggests. Both the search itself and the relevance assessment are very complex, and many different strategies are employed, including the creatively “misuse” of existing tools, since no appropriate tools exist or are unknown to the participants.Conclusion Many of the issues we found relate directly or indirectly to the application of the FAIR principles, but some, like a greater need for dataset search literacy, go beyond that. Both infrastructure and tools offered for dataset search could be tailored more tightly to the observed work processes, particularly by offering more interconnectivity between datasets, literature, and other relevant materials.

Background The engineering of elaborate and innovative tools to navigate the ever growing biomedical knowledge base, instanced in PubMed/Medline, must be guided by genuine case studies addressing `real world´ user needs. Furthermore, algorithm-based predictions regarding `similarity´, `relatedness´ or `relevance´ of pieces of information (e.g. relevance ranking) should be transparent and comprehensible to users. Results We here present a corpus of abstracts (n = 300) annotated on document level representing three case studies in the experimental biomedical domain. The SMAFIRA corpus mirrors `real-world´ information retrieval needs, i.e. the identification of potential alternatives to given animal experiments that support `equivalent´ scientific purposes while using basically different experimental methodology. Since in most cases not even the authors of `relevant´ research papers are aware of such a possible implication of their experimental approaches, our case studies actually illustrate knowledge discovery. Annotation of abstracts (regarding `equivalence´) was conducted by one researcher with broad domain knowledge (in one case study supported by a second opinion from a domain expert) and was informed by a newly created model describing distinguishable stages in experimental biomedicine. Furthermore, such stages were linked to generic scientific purposes. This perspective thus may share some commonalities with topic modelling approaches. Annotation of `relevance´ (i.e. `equivalence´ of scientific purpose plus alternative methodology) relied on expert knowledge in the domain of animal use alternatives. The case studies were used for an evaluation of rankings which were provided by the `similar articles´ algorithm employed in PubMed. Conclusions Building on approved techniques utilized in the domain of intellectual property, we have adapted the concept of `equivalence´ to support a transparent, reproducible and stringent comparison of biomedical textual documents with regards to the implied scientific objectives. This concept may allow for text mining with improved resolution and may aid the retrieval of appropriate animal use alternatives. Computer science researchers in the field of biomedical knowledge discovery may also use our corpus, which is designed to grow essentially in the near future, as a reliable and informative benchmark for the evaluation of algorithms supporting such a goal. Annotations are available from GitHub.

In previous studies on user behavior with Digital Scholarly Editions (DSE), we found that annotating the text is a key technique for working with the text. In this follow-up study, we invited volunteers to perform open research tasks on a DSE by Lope de Vega, providing the annotation tool hypothes.is to support their tasks. What we found is that none of the participants used the tool extensively, yet, it was clear that annotation of text was a major part of their workflow. During a focus discussion after the experiment, the participants gave examples of how they would compensate for a lack of appropriate tools, often at the cost of considerable extra work and overhead. More research is needed to discover why the users were not accepting the tools provided. So, we propose a human-centered structured longitudinal approach to design an annotation tool that would actually be usable.

In this paper, we are analysing the quality of use of digital scholarly editions (DSEs), via usability testing. We do a competitor analyses in which a small sample of target users is gathered in a usability lab, asked to answer research questions via different research tools, and finally to rate the experience. In our first study, we compared three DSEs and found that not all the editions are ideal in their usability and that efficiency is valued more than content. In a second study, we pitched three different media: a printed book, a digitized edition (or PDF) and a digital edition of Nietzsche Also Sprach Zarathustra. We found that users preferred using the PDF edition due to its higher accessibility, portability, and annotation features. This paper introduces our user research and hypothesizes on the benefit and feasibility of developing DSE with a user requirement document.

Protecting animal rights and reducing animal suffering in experimentation is a globally recognized goal in science. Yet numbers have been rising, especially in basic research. While most scientists agree that they would prefer to use less invasive methods, studies have shown that current information systems are not equipped to support the search for alternative methods. In this paper, we outline our investigations into the problem. We look into supervised and semi-supervised methods and outline ways to remedy the problem. We learned that machine assisted methods can identify the documents in question, but they are not perfect yet and in particular the question about gathering sufficient training data is unsolved.

The Thesaurus for the Social Sciences (TheSoz) is a Linked Dataset in SKOS format, which serves as a crucial instrument for information retrieval based on e.g. document indexing or search term recommendation. Thesauri and similar controlled vocabularies build a linking bridge for other datasets from the Linked Open Data cloud - even between different domains. The information and knowledge, which is exposed by such links, can be processed by Semantic Web applications. In this article the conversion process of the TheSoz to SKOS is described including the analysis of the original dataset and its structure, the mapping to adequate SKOS classes and properties, and the technical conversion. Furthermore mappings to other datasets and the appliance of the TheSoz are presented. Finally, limitations and modeling issues encountered during the creation process are discussed., to appear in Semantic Web - Interoperability, Usability, Applicability (IOS Press)

Digital libraries are nowadays expected to contain more than books and articles. All relevant sources of information for a scholar should be available, including research data. However, does literature retrieval work for data sets as well? In the context of a requirement analysis of a data catalogue for quantitative Social Science research data, we tried to find answers to this question. We conducted two user studies with a total of 53 participants and found similarities and important differences in the users’ needs when searching for data sets in comparison to those already known in literature search. In particular, quantity and quality of metadata are far more important in data set search than in literature search, where convenience is most important. In this paper, we present the methodology of these two user studies, their results and challenges for data set retrieval system that can be derived thereof. One of our key findings is that for empirical social scientists, the choice of research data is more relevant than the choice of literature; therefore they are willing to put more effort into the retrieval process. Due to our choice of use case, our initial findings are limited to the field of Social Sciences. However, because of the similar characteristics for data sets also in other research areas, such as Economics, we assume that our results are applicable for them as well.

Mit den XML-basierten Sprachen GML, XSLT und SVG lassen sich Geodaten nicht nur anwendungsorientiert modellieren, sondern auch kartenahnlich visualisieren. In dieser Fallstudie zeigen wir das, indem wir realistische Geodatenbestande der Landesvermessungsamter zunachst mit der Geography Markup Language (GML) nachmodellieren. So mit GML strukturierte Daten werden dann mit der Extensible Stylesheet Language for Transformation (XSLT) auf Elemente der Sprache Scalable Vector Graphics (SVG) abgebildet. Dabei wird der Prozess der kartografischen Visualisierung durch XSLT-Konstrukte modelliert und auch gleichzeitig implementiert. Als Ergebnis erhalten wir Grafiken, die den entsprechenden Karten der Landesamter zumindest nicht unahnlich sind.

Large scale Linked Data is often based on relational databases and thereby tends to be modeled with rich object properties, specifying the exact relationship between two objects, rather than a generic is-a or part-of relationship. We study this phenomenon on government issued statistical data, where a vested interest exists in matching such object properties for data integration. We leverage the fact that while the labeling of the properties is often heterogeneous, e.g. ex1:geo and ex2:location, they link to individuals of semantically similar code lists, e.g. country lists. State-of-the-art ontology matching tools do not use this effect and therefore tend to miss the possible correspondences. We enhance the state-of-the-art matching process by aligning the individuals of such imported ontologies separately and computing the overlap between them to improve the matching of the object properties. The matchers themselves are used as black boxes and are thus interchangeable. The new correspondences found with this method lead to an increase of recall up to 2.5 times on real world data, with only a minor loss in precision.

Citations and their accompanying information such as citation context is a very important consideration when searching for relevant information in literature. With literature continually expanding, obtaining this information in an efficient and effective way, using a standard search tool becomes a very laborious task. In addition, scientific articles found in the literature employ rich interrelationships e.g., citation and their accompanying information making it very difficult for standard search applications to present these interrelationships to the searcher in an efficient manner. In order to alleviate this intensive task, we have designed a tool VisNavi (Visualization and Navigation) that supports scientific researchers in their literature review, in a form of a visualized star-center approach that, for the current treated paper places a citing author at its center and the richly embedded interrelationships are spun around it. The star design enables researchers to gain a clearer insight by interactively exploring these rich interrelationships along with their accompanying information. We designed a human judgment experiment to obtain a human rating on the functionality of the tool. We then cast the human rating as a reference point to improve the tool's design.

The significant increase in content of online social media such as product reviews, blogs, forums etc., have led to an increasing attention to sentiment analysis tools and approaches that make use of mining this substantially growing content. The aim of this paper is to develop a robust classification approach of customer reviews based on a self-annotated domain-specific corpus by applying a statistical approach i.e., mutual information. First, subjective words in each test sentence are identified. Second, ambiguous adjectives such as high, low, large, many etc., are disambiguated based on their accompanying noun using a conditional mutual information approach. Third, a mutual information approach is applied to find the sentiment orientation (polarity) of the identified subjective words based on analyzing their statistical relationship with the manually annotated sentiment labels within a sizeable sentiment training data. Fourth, since negation plays a significant role in flipping the sentiment polarity of an identified sentiment word, we estimate the role of negation in affecting the classification accuracy. Finally, the identified polarity for each test sentence is evaluated against experts' annotation.

When looking at the relationship between two terms, we are not only interested on how much they are related, but how we may explain this relationship to the user. This is an open problem in ontology matching, but also in other tasks, from information retrieval to lexicography. In this paper, we propose a solution based on snippets taken from Wikipedia. These snippets are found by looking for connectors between the two terms, e.g. the terms themselves, but also terms that occur often in both articles or terms that link to both articles. With a user study, we establish that this is particularly useful when dealing with not well known relationships, but well-known concepts. The users were learning more about the relationship and were able to grade it accordingly. On real life data, there are some issues with near synonyms, which are not detected well and terms from different communities, but aside from that we get usable and useful explanations of the term relationships.

The process of retrieving information for literature review purposes differs from traditional web information retrieval. Literature reviews differentiate between the weightiness of the retrieved data segments. For example, citations and their accompanying information, such as cited author, citation context etc., are a very important consideration when searching for relevant information in literature. However, this information is integrated into a scientific paper, in rich interrelationships, making it very complicated for standard search systems to present and track them efficiently. In this paper, we demonstrate a system, VisNavi, in the form of a visualized star-centered approach that introduces the rich citation interrelationships to the searchers in an effective and navigational appearance.

Research data and publications are usually stored in separate and structurally distinct information systems. Often, links between these resources are not explicitly available which complicates the search for previous research. In this paper, we propose a pattern induction method for the detection of study references in full texts. Since these references are not specified in a standardized way and may occur inside a variety of different contexts --- i.e., captions, footnotes, or continuous text --- our algorithm is required to induce very flexible patterns. To overcome the sparse distribution of training instances, we induce patterns iteratively using a bootstrapping approach. We show that our method achieves promising results for the automatic identification of data references and is a first step towards building an integrated information system.

Use Cases Related to an Ontology of the Data Documentation Initiative

With the rise of the open data movement a lot of statistical data has been made publicly available by governments, statistical offices and other organizations. First efforts to visualize are made by the data providers themselves. Data aggregators go a step beyond: they collect data from different open data repositories and make them comparable by providing data sets from different providers and showing different statistics in the same chart. Another approach is to visualize two different indicators in a scatter plot or on a map. The integration of several data sets in one graph can have several drawbacks: different scales and units are mixed, the graph gets visually cluttered and one cannot easily distinguish between different indicators. Our approach marks a combination of (1) the integration of live data from different data sources, (2) presenting different indicators in coordinated visualizations and (3) allows adding user visualizations to enrich official statistics with personal data. Each indicator gets its own visualization, which fits best for the individual indicator in case of visualization type, scale, unit etc. The different visualizations are linked, so that related items can easily be identified by using mouse over effects on data items., To be published in Proceedings of the 5th International Workshop on Web APIs and Services Mashups Proceedings (Mashups '11)

In this paper we present a novel method to connect data on the visualization level. In general, visualizations are a dead end, when it comes to reusability. Yet, users prefer to work with visualizations as evidenced by WYSIWYG editors. To enable users to work with their data in a way that is intuitive to them, we have created Vizgr. Vizgr.com offers basic visualization methods, like graphs, tag clouds, maps and time lines. But unlike normal data visualizations, these can be re-used, connected to each other and to web sites. We offer a simple opportunity to combine diverse data structures, such as geo-locations and networks, with each other by a mouse click. In an evaluation, we found that over 85 % of the participants were able to use and understand this technology without any training or explicit instructions., To be published in Proceedings of 7th International Conference on Web Information Systems and Technologies (WEBIST), 2011

Fun in computer games depends on many factors. While some factors like uniqueness and humor can only be measured by human subjects, in a strategical game, the rule system is an important and measurable factor. Classics like chess and GO have a millennia-old story of success, based on clever rule design. They only have a few rules, are relatively easy to understand, but still they have myriads of possibilities. Testing the deepness of a rule-set is very hard, especially for a rule system as complex as in a classic strategic computer game. It is necessary, though, to ensure prolonged gaming fun.In our approach, we use artificial intelligence (AI) to simulate hours of beta-testing the given rules, tweaking the rules to provide more game-playing fun and deepness. To avoid making the AI a mirror of its programmer's gaming preferences, we not only evolved the AI with a genetic algorithm, but also used three fundamentally different AI paradigms to find boring loopholes, inefficient game mechanisms and, last but not least, complex erroneous behavior.

A neural network trained using the counterpropagation algorithm to produce stochastic conditional simulations is applied and evaluated on a real dataset. This type of network is a non-parametric clustering algorithm not constrained by assumptions (i.e. normal distributions) and is well suited for risk and uncertainty analysis given spatially auto- correlated data. Detailed geophysical measurements from a slab of Berea sandstone are used to allow comparison with a traditional geostatistical method of producing conditional simulations known as sequential Gaussian simulation. Equiprobable simulations and estimated fields of air permeability are generated using an anisotropic spatial structure extracted from a subset of observation data. Results from the counterpropagation network are statistically similar to the geostatistical methods and original reference fields. The combination of simplicity and computational speed make the method ideally suited for environmental subsurface characterization and other earth science applications with spatially auto- correlated variables.

Finding out which genes are expressed in which circumstances is one of the most common tasks in text mining for bioinformatics. But usually the derived data comes from the abstract or other describing texts in the literature. In the age of modern high-throughput microarray analysis, however, there is too much data to be described textual; instead this data often comes in form of tables. In this paper, we are looking specifically at the tables, an approach to our knowledge never described before. The goal is to attach gene names found in tables to their context for a convenient literature review. In order to do so, matching literature has to be downloaded and pre-processed. After that has been done, gene names or protein names can be found through a fast and reliable search, presenting all the associated literature at a glance.

In this paper we present our approach to model and simulate signal transduction pathways as discrete event systems by means of Live Sequence Charts (LSCs) and the Play-Engine tool. To reach this goal we conceptually mapped signal transduction pathways to LSCs and developed an integrated software architecture to generate pathways as discrete event models automatically out of the database.

This paper introduces a new technique of document clustering based on frequent senses. The proposed system, GDClust (graph-based document clustering) works with frequent senses rather than frequent keywords used in traditional text mining techniques. GDClust presents text documents as hierarchical document-graphs and utilizes an apriori paradigm to find the frequent subgraphs, which reflect frequent senses. Discovered frequent subgraphs are then utilized to generate sense-based document clusters. We propose a novel multilevel Gaussian minimum support approach for candidate subgraph generation. GDClust utilizes English language ontology to construct document-graphs and exploits graph-based data mining technique for sense discovery and clustering. It is an automated system and requires minimal human interaction for the clustering purpose.

Connecting scientific databases is a challenging task which can be supported by ontologies describing them on a semantical level. Unfortunately, ontologies for databases are rarely used, because schemas of research databases change rapidly while the ontologies grow as well. Maintaining the connection between both is time consuming manual work. Therefore it is worthwhile to reduce the required work by automating tasks. We propose an approach that allows the database schema and the ontology to change and evolve, without ever losing their connection to each other. This is done by mapping database schemas to ontologies, enriching these ontologies with semantical information and transfering schema changes to the ontology.

This paper demonstrates the first steps of an automation process to develop models of signal transduction pathways using discrete modelling languages. The whole approach consists of modelling, validation, animation, linking databases to simulation tools and also the qualitative analysis of the data. In this paper, we detail the modelling and simulation of the TLR4 pathway with a coloured petri net simulation tool and the validation of this model against the semantic and mechanistic map from a biological database. These graphical maps contain all necessary reactions as a figure. We start with an UML class diagram to understand the static structure of molecules involved in the TLR4 pathway. Afterwards we model and simulate each "pathway step reaction" - one after another - to get the behaviour of the final system. The result is a model of the pathway which can be used in simulations, derived solely from basic chemical reactions in the database. Also, it is a lesson on critical points where human decision-making is needed, because not all the required information is stored directly in the database.

Currently, knowledge from biological research is stored in hundreds of databases, counting only public accessible ones. Finding specific data in these is a challenging task which can be supported by ontologies describing them. The maintenance of a corresponding ontology is time consuming manual work, because research database schemas change rapidly. Our project will reduce the work by automating tasks, like a generation process and applying schema changes to the corresponding ontology. We call the proposed method coevolution, because database schema and ontology are allowed to evolve independently without ever losing their connection to each other. Our method consists of initial ontology generation, manual annotation and change propagation steps.

The aim of literature retrieval is to find significant papers on a given topic. In previous publications, we examined the use of choosing these papers based on the pictures they include. To refine this approach we seek to employ picture classification to further narrow down the number of interesting pictures presented. This can be useful, for example, when looking for the results of specific experiments. The classification can also be useful as a data cleansing step, to omit all unnecessary pictures not used as a figure. We use a method originally designed to distinguish between photos and computer-generated pictures on the web. We show that this method can not only be used to distinguish between raw data and derived representation figures, we can also reliably eliminate non-figure pictures in the document, like text pages and logos. We tested this approach on two different data sets with different topics and different non-figure problems, both with satisfactory results.

Ontologies are one of the key technologies for data integration and meta-databases, by connecting databases on a semantical level. Still, everything fails when one of the database schemas changes: specific parts of the ontology have to be reconstructed by hand. We propose an approach that allows the database schema and the ontology to change and evolve, without ever losing their connection to each other. We call that "coevolution". Coevolution cannot be completely automated, as data definition languages do not define the change of semantical concepts, but only technical schema changes. In this paper we present our mapping of database schemas to ontologies, describe how these ontologies can be enriched by semantical information and show our approach to transfer schema changes to the ontology

In this paper we present a picture search engine for life science literature and show how it can be used to improve literature preselection. This preselection is needed as a way to compensate for the vast amounts of literature that are available. While searching for DNA binding sites for example, we wanted to add the results of specific experiments (DNAse I footprint and EMSA) to our database. The preselection via abstract search was very unspecific (150 000 hits), but by looking for paper with images concerning the experiments, we could improve precision immensely. They are displayed like hits in a search engine, allowing easy and quick quality assessment without having to read through the whole paper. The images are found by their annotation in the paper: the figure caption. To identify that, we analyse the layout of the paper: the position of the image and the surrounding text.

Domain ontologies and XML Schemas serve to describe domain data models although they follow different modelling goals. By lifting the syntactic level of XML documents and validating XML Schemas to the semantic level of OWL ontologies and their RDF representations in an automatic way, all the information located in the XML Schemas of the domains can be reused by ontology engineers and domain experts to design domain ontologies from scratch. As this approach supports all components of the XML Schema metamodel, it is ensured that unexceptionally any XML Schema can be converted into a generated ontology. As structures of generated ontologies might be quite complex, domain ontologies can be inferred automatically by means of SWRL rules. Saved time and effort can then be used to add domain-specific semantic information, not covered by underlying XML Schemas, to the domain ontologies.

Ontologies are one of the key technologies for data integration and meta-databases, by connecting databases at a semantical level. Still, the database has to be connected to the ontology and vice versa. To achieve this, we propose a two step process. First, we automatically generate an ontology directly from the database schema. Next, we annotate this with concepts from a domain specific ontology. We demonstrate this for signal transduction pathways, which describe how cells can react to external stimuli. In this paper we present our mapping of database schemas to ontologies and describe how these ontologies can be enriched by semantical information.