Background and foreground knowledge in dynamic ontology construction

Ontologies have become a key component in the Semantic Web and Knowledge management. One accepted goal is to construct ontologies from a domain specific set of texts. An ontology reflects the background knowledge used in writing and reading a text. However, a text is an act of knowledge maintenance, in that it re-enforces the background assumptions, alters links and associations in the ontology, and adds new concepts. This means that background knowledge is rarely expressed in a machine interpretable manner. When it is, it is usually in the conceptual boundaries of the domain, e.g. in textbooks or when ideas are borrowed into other domains. We argue that a partial solution to this lies in searching external resources such as specialized glossaries and the internet. We show that a random selection of concept pairs from the Gene Ontology do not occur in a relevant corpus of texts from the journal Nature. In contrast, a significant proportion can be found on the internet. Thus, we conclude that sources external to the domain corpus are necessary for the automatic construction of ontologies.

Ontologies, taxonomies, thesauri:learning from texts

The use of ontologies as representations of knowledge is widespread but their construction, until recently, has been entirely manual. We argue in this paper for the use of text corpora and automated natural language processing methods for the construction of ontologies. We delineate the challenges and present criteria for the selection of appropriate methods. We distinguish three ma jor steps in ontology building: associating terms, constructing hierarchies and labelling relations. A number of methods are presented for these purposes but we conclude that the issue of data-sparsity still is a ma jor challenge. We argue for the use of resources external tot he domain specific corpus.

miniDVMS v1.8 : A user manual(The data visualization and modeling system)

Today, the data available to tackle many scientific challenges is vast in quantity and diverse in nature. The exploration of heterogeneous information spaces requires suitable mining algorithms as well as effective visual interfaces. miniDVMS v1.8 provides a flexible visual data mining framework which combines advanced projection algorithms developed in the machine learning domain and visual techniques developed in the information visualisation domain. The advantage of this interface is that the user is directly involved in the data mining process. Principled projection methods, such as generative topographic mapping (GTM) and hierarchical GTM (HGTM), are integrated with powerful visual techniques, such as magnification factors, directional curvatures, parallel coordinates, and user interaction facilities, to provide this integrated visual data mining framework. The software also supports conventional visualisation techniques such as principal component analysis (PCA), Neuroscale, and PhiVis. This user manual gives an overview of the purpose of the software tool, highlights some of the issues to be taken care while creating a new model, and provides information about how to install and use the tool. The user manual does not require the readers to have familiarity with the algorithms it implements. Basic computing skills are enough to operate the software.

Multiple sclerosis: natural history, diagnosis and changes in the brain

This article describes the general symptoms, diagnosis and changes in the nervous system in multiple sclerosis. A second article will describe the specific visual symptoms which are believed to be characteristic of the disease.

Probabilistic methods for drug discovery

This thesis introduces a flexible visual data exploration framework which combines advanced projection algorithms from the machine learning domain with visual representation techniques developed in the information visualisation domain to help a user to explore and understand effectively large multi-dimensional datasets. The advantage of such a framework to other techniques currently available to the domain experts is that the user is directly involved in the data mining process and advanced machine learning algorithms are employed for better projection. A hierarchical visualisation model guided by a domain expert allows them to obtain an informed segmentation of the input space. Two other components of this thesis exploit properties of these principled probabilistic projection algorithms to develop a guided mixture of local experts algorithm which provides robust prediction and a model to estimate feature saliency simultaneously with the training of a projection algorithm.Local models are useful since a single global model cannot capture the full variability of a heterogeneous data space such as the chemical space. Probabilistic hierarchical visualisation techniques provide an effective soft segmentation of an input space by a visualisation hierarchy whose leaf nodes represent different regions of the input space. We use this soft segmentation to develop a guided mixture of local experts (GME) algorithm which is appropriate for the heterogeneous datasets found in chemoinformatics problems. Moreover, in this approach the domain experts are more involved in the model development process which is suitable for an intuition and domain knowledge driven task such as drug discovery. We also derive a generative topographic mapping (GTM) based data visualisation approach which estimates feature saliency simultaneously with the training of a visualisation model.

Investigating the effects of antiepileptic drugs on the electrophysiology of vision

The principal aim of this work was to examine the effects of antiepileptic drugs (AEDs) on vision. Vigabatrin acts by increasing GABA at brain inhibitory synapses by irreversibly binding to GABA-transaminase. Remacemide is a novel non-competitive NMDA receptor antagonist and fast sodium channel inhibitor that results in the inhibition of the NMDA receptors located in the neuronal membrane calcium channels increasing glutamate in the brain. Vigabatrin has been shown to cause a specific pattern of visual field loss, as one in three adults taking vigabatrin have shown a bilateral concentric constriction. Remacemide has unknown effects on vision. The majority of studies of the effects of AEDs on vision have not included the paediatric population due to difficulties assessing visual field function using standard perimetry testing. Evidently an alternative test is required to establish and monitor visual field problems associated with AEDs both in children and in adults who cannot comply with perimetry. In order to test paediatric patients exposed to vigabatrin, a field-specific visual evoked potential was developed. Other tests performed on patients taking either vigabatrin or remacemide were electroretinograms, electro-oculograms, multifocal VEPs and perimetry. Comparing these tests to perimetry results from vigabatrin patients the field specific VEP was found to have a high sensitivity and specificity, as did the 30Hz flicker amplitude. The modified VEP was also found to provide useful results in vigabatrin patients. Remacemide did not produce a similar visual field loss to vigabatrin although macular vision was affected. The field specific VEP is a useful method for detecting vigabatrin associated visual field loss that is well tolerated by young children. This technique combined with the ERG under light adapted (30Hz flicker) condition is presently the superior method for detecting vigabatrin-attributed peripheral field defects present in children below the developmental age of 9. The effects of AEDs on vision should be monitored carefully and the use of multifocal stimulation allows for specific areas of the retina and visual pathway to be monitored.

Self-training from labeled features for sentiment analysis

Sentiment analysis concerns about automatically identifying sentiment or opinion expressed in a given piece of text. Most prior work either use prior lexical knowledge defined as sentiment polarity of words or view the task as a text classification problem and rely on labeled corpora to train a sentiment classifier. While lexicon-based approaches do not adapt well to different domains, corpus-based approaches require expensive manual annotation effort. In this paper, we propose a novel framework where an initial classifier is learned by incorporating prior information extracted from an existing sentiment lexicon with preferences on expectations of sentiment labels of those lexicon words being expressed using generalized expectation criteria. Documents classified with high confidence are then used as pseudo-labeled examples for automatical domain-specific feature acquisition. The word-class distributions of such self-learned features are estimated from the pseudo-labeled examples and are used to train another classifier by constraining the model's predictions on unlabeled instances. Experiments on both the movie-review data and the multi-domain sentiment dataset show that our approach attains comparable or better performance than existing weakly-supervised sentiment classification methods despite using no labeled documents.

Learning sentiment classification model from labeled features

We propose a novel framework where an initial classifier is learned by incorporating prior information extracted from an existing sentiment lexicon. Preferences on expectations of sentiment labels of those lexicon words are expressed using generalized expectation criteria. Documents classified with high confidence are then used as pseudo-labeled examples for automatical domain-specific feature acquisition. The word-class distributions of such self-learned features are estimated from the pseudo-labeled examples and are used to train another classifier by constraining the model's predictions on unlabeled instances. Experiments on both the movie review data and the multi-domain sentiment dataset show that our approach attains comparable or better performance than exiting weakly-supervised sentiment classification methods despite using no labeled documents.

Case based design of knitwear

In the developed world we are surrounded by man-made objects, but most people give little thought to the complex processes needed for their design. The design of hand knitting is complex because much of the domain knowledge is tacit. The objective of this thesis is to devise a methodology to help designers to work within design constraints, whilst facilitating creativity. A hybrid solution including computer aided design (CAD) and case based reasoning (CBR) is proposed. The CAD system creates designs using domain-specific rules and these designs are employed for initial seeding of the case base and the management of constraints. CBR reuses the designer's previous experience. The key aspects in the CBR system are measuring the similarity of cases and adapting past solutions to the current problem. Similarity is measured by asking the user to rank the importance of features; the ranks are then used to calculate weights for an algorithm which compares the specifications of designs. A novel adaptation operator called rule difference replay (RDR) is created. When the specifications to a new design is presented, the CAD program uses it to construct a design constituting an approximate solution. The most similar design from the case-base is then retrieved and RDR replays the changes previously made to the retrieved design on the new solution. A measure of solution similarity that can validate subjective success scores is created. Specification similarity can be used as a guide whether to invoke CBR, in a hybrid CAD-CBR system. If the newly resulted design is suffciently similar to a previous design, then CBR is invoked; otherwise CAD is used. The application of RDR to knitwear design has demonstrated the flexibility to overcome deficiencies in rules that try to automate creativity, and has the potential to be applied to other domains such as interior design.

The impact of general intellectual ability and white matter volume on the functional outcome of patients with bipolar disorder and their relatives

Background: There is substantial evidence that cognitive deficits and brain structural abnormalities are present in patients with Bipolar Disorder (BD) and in their first-degree relatives. Previous studies have demonstrated associations between cognition and functional outcome in BD patients but have not examined the role of brain morphological changes. Similarly, the functional impact of either cognition or brain morphology in relatives remains unknown. Therefore we focused on delineating the relationship between psychosocial functioning, cognition and brain structure, in relation to disease expression and genetic risk for BD. Methods: Clinical, cognitive and brain structural measures were obtained from 41 euthymic BD patients and 50 of their unaffected first-degree relatives. Psychosocial function was evaluated using the General Assessment of Functioning (GAF) scale. We examined the relationship between level of functioning and general intellectual ability (IQ), memory, attention, executive functioning, symptomatology, illness course and total gray matter, white matter and cerebrospinal fluid volumes. Limitations: Cross-sectional design. Results: Multiple regression analyses revealed that IQ, total white matter volume and a predominantly depressive illness course were independently associated with functional outcome in BD patients, but not in their relatives, and accounted for a substantial proportion (53%) of the variance in patients' GAF scores. There were no significant domain-specific associations between cognition and outcome after consideration of IQ. Conclusions: Our results emphasise the role of IQ and white matter integrity in relation to outcome in BD and carry significant implications for treatment interventions. © 2010 Elsevier B.V.

The role of working memory in compulsive checking and OCD:a systematic classification of 58 experimental findings

Despite the large body of research regarding the role of memory in OCD, the results are described as mixed at best (Hermans et al., 2008). For example, inconsistent findings have been reported with respect to basic capacity, intact verbal, and generally affected visuospatial memory. We suggest that this is due to the traditional pursuit of OCD memory impairment as one of the general capacity and/or domain specificity (visuospatial vs. verbal). In contrast, we conclude from our experiments (i.e., Harkin & Kessler, 2009, 2011; Harkin, Rutherford, & Kessler, 2011) and recent literature (e.g., Greisberg & McKay, 2003) that OCD memory impairment is secondary to executive dysfunction, and more specifically we identify three common factors (EBL: Executive-functioning efficiency, Binding complexity, and memory Load) that we generalize to 58 experimental findings from 46 OCD memory studies. As a result we explain otherwise inconsistent research – e.g., intact vs. deficient verbal memory – that are difficult to reconcile within a capacity or domain specific perspective. We conclude by discussing the relationship between our account and others', which in most cases is complementary rather than contradictory.

Self-explanation in adaptive systems based on runtime goal-based models

The behaviour of self adaptive systems can be emergent, which means that the system’s behaviour may be seen as unexpected by its customers and its developers. Therefore, a self-adaptive system needs to garner confidence in its customers and it also needs to resolve any surprise on the part of the developer during testing and maintenance. We believe that these two functions can only be achieved if a self-adaptive system is also capable of self-explanation. We argue a self-adaptive system’s behaviour needs to be explained in terms of satisfaction of its requirements. Since self-adaptive system requirements may themselves be emergent, we propose the use of goal-based requirements models at runtime to offer self-explanation of how a system is meeting its requirements. We demonstrate the analysis of run-time requirements models to yield a self-explanation codified in a domain specific language, and discuss possible future work.

webComputing Service Framework

Presented is webComputing – a general framework of mathematically oriented services including remote access to hardware and software resources for mathematical computations, and web interface to dynamic interactive computations and visualization in a diversity of contexts: mathematical research and engineering, computer-aided mathematical/technical education and distance learning. webComputing builds on the innovative webMathematica technology connecting technical computing system Mathematica to a web server and providing tools for building dynamic and interactive web-interface to Mathematica-based functionality. Discussed are the conception and some of the major components of webComputing service: Scientific Visualization, Domain- Specific Computations, Interactive Education, and Authoring of Interactive Pages.

MPEG-7 Based Image Retrieval on the World Wide Web

Due to the rapid growth of the number of digital media elements like image, video, audio, graphics on Internet, there is an increasing demand for effective search and retrieval techniques. Recently, many search engines have made image search as an option like Google, AlltheWeb, AltaVista, Freenet. In addition to this, Ditto, Picsearch, can search only the images on Internet. There are also other domain specific search engines available for graphics and clip art, audio, video, educational images, artwork, stock photos, science and nature [www.faganfinder.com/img]. These entire search engines are directory based. They crawls the entire Internet and index all the images in certain categories. They do not display the images in any particular order with respect to the time and context. With the availability of MPEG-7, a standard for describing multimedia content, it is now possible to store the images with its metadata in a structured format. This helps in searching and retrieving the images. The MPEG-7 standard uses XML to describe the content of multimedia information objects. These objects will have metadata information in the form of MPEG-7 or any other similar format associated with them. It can be used in different ways to search the objects. In this paper we propose a system, which can do content based image retrieval on the World Wide Web. It displays the result in user-defined order.

One Approach to Knowledge Mapping for International Student Portal

* The work is partly supported by RFFI grant 08-07-00062-a