Combined Weighting Research Articles

Enriching content-based image retrieval with multilingual search terms

Summary Content-based visual image retrieval is a research domain with the goal to allow efficient access to the large amount of visual information that is being produced in medical institutions. Currently, visual retrieval is most often strictly separated from textual information extraction and retrieval in medical records. The complementary nature of the two methods by contrast invites to use the two together in an integrated fashion. We use a visual retrieval system (medGIFT) and a textual search engine powered with biomedical terminological resources (easyIR) together on a data set presented at the imageCLEF image retrieval competition. Both systems are available free of charge as open source. The dataset is also publicly available to make results reproducible and comparable. Results show that a simple combination of visual and textual features for retrieval improves performance significantly for fully automatic retrieval as well as for runs with manual relevance feedback. The currently applied techniques are fairly simple combinations and better results can be expected when optimizing the combined weighting based on learning data. Visual and textual features should be used together for information retrieval whenever they are both available to allow optimal access to varied data sources.

Marking the Difference: An investigation of the criteria used for assessing undergraduate dissertations in a business school

Although the outcomes of a first degree are complex, UK universities continue to use a single measure for grading the student's performance—namely the degree classification. This classification has significant implications for the student, prospective employers, and for the awarding institution. There is some evidence of changing perceptions of degree classes, for example that third class (and even, in some cases, lower second class) awards are unacceptable. At the same time there is concern about maintenance of quality standards in HE, and of possible grade inflation. The final degree class, although a single measure, results from an accumulation of marks achieved throughout the degree course; it is the combined weighting of individual tutor marks which ultimately determines the student's degree class. This study focuses on the marking process, and the application of criteria within one final year module—the dissertation. The dissertation formed a significant part of the final degree weighting, and there was evidence of the use of different criteria and approaches to marking by staff within the faculty. The study employed a qualitative research methodology—including: a structured group discussion where the issue of criteria was raised and brainstormed; and a questionnaire to staff which asked open questions about the criteria which each used when marking dissertations. The responses were analysed, and common ground and differences identified. Questions are raised concerning (a) the tension which surfaces within the marking process, between individual tutor freedom and institutional demands, and (b) staff development opportunities within the marking process.

A method of optimal scaling for multivariate ordinal data and its extensions

This paper develops a method of optimal scaling for multivariate ordinal data, in the framework of a generalized principal component analysis. This method yields a multidimensional configuration of items, a unidimensional scale of category weights for each item and, optionally, a multidimensional configuration of subjects. The computation is performed by alternately solving an eigenvalue problem and executing a quasi-Newton projection method. The algorithm is extended for analysis of data with mixed measurement levels or for analysis with a combined weighting of items. Numerical examples and simulations are provided. The algorithm is discussed and compared with some related methods.

Agents and Recipient Objects in the Development of Early Symbolic Play

The development of 2 components of early symbolic play--children's symbolic transformations of agents of action and recipient objects of action--was assessed in 48 children from 14 to 25 months of age. A 5-step sequence was predicted based on a combined weighting of agent substitutions and recipient object substitutions in each step. Children observed an adult model who pretended to talk into a telephone using each agent and object substitution in all combinations and then were allowed to demonstrate their own imitative pretending for each step. Based on the children's profiles of steps passed and failed, the sequence was found to form a Guttman scale and be age related. 2 types of possible underlying sequences were discussed.

S.A.W. device reflectors weighted by combination of withdrawn and segmented lines

A technique for weighting surface acoustic wave (s.a.w.) device reflectors which combines withdrawal weighting and segmented lines is described. This scheme has the advantage of lower photolithographic mask costs through the partial use of withdrawal weighting, and the advantage of closely approximating the desired weighting function by partial use of segmented (dotted) lines. Effectiveness of this combined weighting technique is demonstrated by its use in sidelobe suppression in two-port s.a.w. resonators; however, combined-type weighted reflectors can be used to advantage in all types of filters utilising reflective arrays.