Sulieman Bani-Ahmad

Context-based literature digital collection search

We identify two issues with searching literature digital collections within digital libraries: (a) there are no effective paper-scoring and ranking mechanisms. Without a scoring and ranking system, users are often forced to scan a large and diverse set of publications listed as search results and potentially miss the important ones. (b) Topic diffusion is a common problem: publications returned by a keyword-based search query often fall into multiple topic areas, not all of which are of interest to users. This paper proposes a new literature digital collection search paradigm that effectively ranks search outputs, while controlling the diversity of keyword-based search query output topics. Our approach is as follows. First, during pre-querying, publications are assigned into pre-specified ontology-based contexts, and query-independent context scores are attached to papers with respect to the assigned contexts. When a query is posed, relevant contexts are selected, search is performed within the selected contexts, context scores of publications are revised into relevancy scores with respect to the query at hand and the context that they are in, and query outputs are ranked within each relevant context. This way, we (1) minimize query output topic diversity, (2) reduce query output size, (3) decrease user time spent scanning query results, and (4) increase query output ranking accuracy. Using genomics-oriented PubMed publications as the testbed and Gene Ontology terms as contexts, our experiments indicate that the proposed context-based search approach produces search results with up to 50% higher precision, and reduces the query output size by up to 70%.

Evaluating score and publication similarity functions in digital libraries

Digital libraries do not assign importance/relevance scores to their publications, authors, or publication venues, even though scores are potentially useful for (a) providing comparative assessment, or “importances”, of publications, authors, publication venues, (b) ranking publications returned in search outputs, and (c) using scores in locating similar publications. Using social networks and bibliometrics, one can define several score functions.

Improved publication scores for online digital libraries via research pyramids

Ranking publications of Online Digital Libraries (ODLs) is useful for (i) providing comparative assessment of publications and (ii) listing relevant ODL search results first in search outputs, enabling users to aggregate pertinent results quickly and easily. Studies show that effective citation-based scoring functions, namely, PageRank, HITS and Citation Count, are highly skewed, and have accuracy problems, possibly due to topic diffusion. In this paper, based on the notion of research pyramids, we propose an a priori technique to assign more effective publication scores. Using the ACM SIGMOD Anthology ODL as a testbed, we show that our approach provides more accurate and less skewed publication scores.

Evaluating Different Ranking Functions for Context-Based Literature Search

Context-based literature digital library search is a new search paradigm that creates an effective ranking of query outputs by controlling query output topic diversity. We define contexts as pre-specified ontology-based terms and locate the paper set of a context based on semantic properties of the context (ontology) term. In order to provide a comparative assessment of papers in a context and effectively rank papers returned as search outputs, prestige scores are attached to all papers with respect to their assigned contexts. In this paper, we present three different prestige score (ranking) functions for the context-based environment, namely, citation-based, text-based, and pattern-based score functions. Using biomedical publications as the test case and Gene Ontology as the context hierarchy, we have evaluated the proposed ranking functions in terms of their accuracy and separability. We have found that text-based and pattern-based score functions yield better accuracy and separability than citation-based score functions.