Large-scale semantic exploration of scientific literature using topic-based hashing algorithms

Abstract

Searching for similar documents and exploring major themes covered across groups of documents are common activities when browsing collections of scientific papers. This manual knowledge-intensive task can become less tedious and even lead to unexpected relevant findings if unsupervised algorithms are applied to help researchers. Most text mining algorithms represent documents in a common feature space that abstract them away from the specific sequence of words used in them. Probabilistic Topic Models reduce that feature space by annotating documents with thematic information. Over this low-dimensional latent space some locality-sensitive hashing algorithms have been proposed to perform document similarity search. However, thematic information gets hidden behind hash codes, preventing thematic exploration and limiting the explanatory capability of topics to justify content-based similarities. This paper presents a novel hashing algorithm based on approximate nearest-neighbor techniques that uses hierarchical sets of topics as hash codes. It not only performs efficient similarity searches, but also allows extending those queries with thematic restrictions explaining the similarity score from the most relevant topics. Extensive evaluations on both scientific and industrial text datasets validate the proposed algorithm in terms of accuracy and efficiency.

PageRank Topic Finder based Algorithm for Multimedia Resources in Preschool Education

“…Most of these problems are reflected in older kindergarten teachers and township kindergartens with relatively backward economic levels. Some kindergarten teachers lack the understanding of multimedia teaching and do not know the important role of multimedia teaching in early childhood teaching, which affects the application effect of multimedia teaching in preschool education [ 10 – 12 ]. Thematic search engine appears in such a background.…”

Scalable Cross-lingual Document Similarity through Language-specific Concept Hierarchies

“…The notion of topics is discarded and therefore the ability to make thematic explorations of documents. Recently, a hashing algorithm that groups similar documents and preserves the notion of topics has been proposed [2]. It defines a hierarchical set-type data where each level of the hierarchy indicates the importance of the topic according to its distribution.…”

“…Documents represented as a weighted mixture of latent topics (per-document topic distributions) are then annotated in these feature spaces with the relation between topics inside each hierarchy level. Regardless of their language, they are then described by cross-lingual concepts (based on WordNet-synset annotations) and hash codes are calculated to summarize their content [2]. The hash expression sets a 3-level hierarchy of cross-lingual concepts.…”

“…More specifically, the similarity metric used to compare the hash codes created from set of topics is the sum of the Jaccard distances for each hierarchy level, i.e. for each set of topics [2]:…”

Editorial: Special issue on Semantic eScience: Methods, tools and applications

“…After a careful reviewing process, two papers were accepted, which focus on discoverability and findability of scientific products. In [2] the authors ease exploration of scientific literature by grouping similar papers according to their topic; while in [3] the authors assist other researchers in describing and summarizing their scientific experiments based on publicly available metadata.…”