Search

Discovering Transcription Factor Binding Sites (TFBS) has immense significance in terms of developing techniques and evaluating regulatory processes in biological systems. The DNA gene sequence encompasses large volume of datasets so a new methodology is needed to analyze them in the quickest possible time. Over the past decades, the planted (ℓ, d) motif discovery methodology has been used for locating TFBS in the genetic region. This paper focuses on developing a new approach for motif identification using planted (ℓ, d) motif discovery algorithm. The proposed algorithm is named ESMD (Emerging Substring based Motif Detection), which is based on two processes: Mining and Combining Emerging Substrings. In the mining step, an array is initially created, based on the suffix array (SA) and the longest common prefix array (LCP). A MapReduce programming model handles the mining of emerging substring process since DNA gene sequences constitute huge data. The next step combines the emerging substrings of different lengths. The resulting models have been evaluated using two different metrics, the Pearson Correlation Coefficient (PCC) and the Area Under Curve (AUC). Both have produced much better results than existing methods. [ABSTRACT FROM AUTHOR]

An introduction to articles published within the issue is presented, including one on a study of the long-term generation capacity investment problem of an independent power generation company, another which applies computational intelligence (CI) techniques for sentiment analysis of electronic commerce applications, and one on the development of an incremental crawler associated with the normal crawl architecture for both surface web and deep web.

This paper studies efficient processing of durable top-k queries on historical time series databases. Durable top-k queries, obtained as an extension of snapshot top-k queries during a certain time period, play a key role in finding objects with durable quality and predicting the status of these objects for successive time intervals by updating the query interval at all timestamps. Web crawling and indexing are tremendously significant in recent times, especially in terms of achieving efficient durable top-k queries from vast quantum of web documents. Existing algorithms that have been employed throw up results that are less than applicable to analyzers. This paper chiefly focuses on web crawling and indexing query terms under their respective categories and updating rank changes at every time interval. Links are crawled using the modified depth-first search (MDFS) algorithm, accessed, and metadata such as the title, keywords, and descriptions extracted. To handle query indexing, novel indexing techniques are proposed to yield efficient results. This study is invaluable for analysts working on large data obtained as a result of crawling and indexing, effectively decreasing their workload. [ABSTRACT FROM AUTHOR]

Nowadays XML based big bibliographic datasets are common in different domains which provide meta data about articles published in that domain. They have well defined tags which give details of the year, title, authors, abstract, keywords, the type of article, the venue of publishing the article and other such specific details about each article. A lot of statistics can be extracted from this dataset. Most of the time the tag pertaining to domain sub topic information associated with the article will be absent in the dataset as it is not an article attribute. Hence for such statistics articles must be mapped to its associated sub domain. This paper investigates this problem and proposes a fast approach to find trending articles and hot topics from XML based big bibliographic datasets. The proposed framework uses domain ontology to first classify articles into its sub topics. Fast detection of hot topics, trending keywords and articles is achieved using novel Map Reduce algorithms implemented on a hadoop distributed framework. Performance comparison demonstrates that it outperforms its non-Map Reduce counterpart in quickly sorting out the trending keywords and titles in a particular hot topic from XML based bibliographic dataset. [ABSTRACT FROM AUTHOR]

Discriminating between the text and non text regions of an image is a complex and challenging task. In contrast to Caption text, Scene text can have any orientation and may be distorted by the perspective projection. Moreover, it is often affected by variations in scene and camera parameters such as illumination, focus, etc. These variations make the design of unified text extraction from various kinds of images extremely difficult. This paper proposes a statistical unified approach for the extraction of text from hybrid textual images (both Scene text and Caption text in an image) and Document images with variations in text by using carefully selected features with the help of multi level feature priority (MLFP) algorithm. The selected features are combinedly found to be the good choice of feature vectors and have the efficacy to discriminate between text and non text regions for Scene text, Caption text and Document images and the proposed system is robust to illumination, transformation/perspective projection, font size and radially changing/angular text. MLFP feature selection algorithm is evaluated with three common ML algorithms: a decision tree inducer (C4.5), a naive Bayes classifier, and an instance based K-nearest neighbour learner and effectiveness of MLFP is shown by comparing with three feature selection methods with benchmark dataset. The proposed text extraction system is compared with the Edge based method, Connected component method and Texture based method and shown encouraging result and finds its major application in preprocessing for optical character recognition technique and multimedia processing, mobile robot navigation, vehicle license detection and recognition, page segmentation and text-based image indexing, etc. [ABSTRACT FROM AUTHOR]

Interactivity, group learning and student engagement are accepted as key features of social constructivist learning theories. The challenge is to understand the interplay between such features in different learning environments. This study focused on the qualitative differences between two interventions—small-groups and whole-class discussions. In both interventions, three short video slices on the abstract topic ‘the physics of superconductivity’ were interspersed with the different discussion styles. The video slices are based on the Bruner stages. Twenty-nine first year university physics students completed a pre-test, underwent the intervention and completed a post-test. The remainder of the data were collected from student drawings, video recordings, observer notes and facilitator feedback. Results indicate that the use of the video slices in both interventions were successful in changing students’ understandings of superconductivity. However, the small groups treatment tended to facilitate questioning, meaning-making and subsequent changes of ideas more so than the whole class discussions. Implications for research and practice are discussed. [ABSTRACT FROM AUTHOR]

We report on a study of what happened when we recycled a conceptual examination question in a first-year university physics course. The question, which was used for three consecutive years, asked about an astronaut's experience of weighing in an orbiting space-craft. Our original intention was to use a phenomenographic approach to look for differences in students' descriptive answers. Having done that, we decided to add a study of the marks that were awarded to those answers. The first time that the question was re-used, the distribution of answers amongst our phenomenographic categories showed a decrease in the common conception that gravity is zero in the satellite and an increase in explanations in terms of free fall. When the question was re-used a second time, that difference was maintained but it was not significantly increased. The distribution of marks for the question was different over the three years in a way that appears to be unrelated to differences in students' conceptual understandings. Differences in the distribution of marks are more likely to be related to differences in marking procedures. We conclude that studies like this one have the potential to contribute to improvements in university assessment procedures. In particular we propose that phenomenographic analysis could be used in the design of marking schemes. [ABSTRACT FROM AUTHOR]

We report on an investigation of students' ideas about gravity after a semester of instruction in physics at university. There are two aspects to the study which was concerned with students' answers to a carefully designed qualitative examination question on gravity. The first aspect is a classification of the answers and a comparative study of the ways the problem was tackled by two large groups of students who had different backgrounds in physics and were exposed to different teaching styles. The second aspect is to investigate how students link concepts to solve the problem. We used a phenomenographic analysis of student responses to extract patterns of reasoning and alternative conceptions behind the solutions. We found no differences between the classes of answers given by students in the two courses. Our analysis also identifies a hierarchy in the complexity of the hypothetical reasoning pathways, which we interpret as reflecting the ways in which students may link concepts and resolve conflicts as they solve the problem. The hypothetical reasoning pathways may help educators to develop instructional material or lecture room dialogue in order to tease out key issues. An unexpected finding is that there is a discrepancy between our conclusion that the two groups of answers are similar and the distribution of marks awarded by the examiner – which implies that the quality of the answers is different for the two groups. [ABSTRACT FROM AUTHOR]

Automatic annotation of images is considered to be an important research problem in image retrieval. Traditional methods are computationally complex and fail to annotate correctly when the number of image classes is large and related. This paper proposes a novel approach, an autoencoder hashing, to categorize images of large-scale image classes. The intra bin classifiers are trained to classify the query image, and the tag weight and tag frequency are computed to achieve a more effective annotation of the query image. The proposed approach has been compared with other existing approaches in the literature using performance measures, such as precision, accuracy, mean average precision (MAP), and F1 score. The experimental results indicate that our proposed approach outperforms the existing approaches. [ABSTRACT FROM AUTHOR]