English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Automatic document summarization has become increasingly important in the age of big data. However, traditional summarization systems have two major weaknesses: term-vector data sparsity and semantic information deficiency. In this paper, we solve these problems by adopting word and sentence embeddings, whose distributed nature enables the relation establishment between words and sentences. Experimental result shows that our models outperform two state-of-the-art graph based summarization systems. Introduction Document summarization is an automatic process of constructing a simple and coherent summary. A good summary should retain most important contents of original document and present them in shorter text as much as possible. As summary greatly shortens time of reading, it is becoming increasingly significant in the fast-growing information age. Document summarization is first introduced by Luhn[1] and have received growing interest among computational linguistics. There are two types of summarizations: abstractive summarization and extractive summarization. In this paper we focus on producing fully automated single-document extractive summary. The pioneer work of Luhn[1] simply used word frequency driven method. Each sentence is represented as a real number, which indicates howmany important words the sentence includes. Dunning[2] improved Luhn’s work by applying log-likelihood ratio test. Edmundson[3] proposed that not only important content words should be considered, but also three other features: cue words, title and heading words. His work has guided many machine learning based algorithms. Lexical chain based summarization[4] concentrates on solving the soft spot that the previous work did not pay attention to words semantic relations. The main strategy is trying to capture similarity between words using man-made lexical resources. However, all the techniques presented above suffer from their own problems. Frequency based methods rely on shallow formal clues in the text. Machine learning algorithms require human generated summaries in their training session. Lexical chain system[4] heavily rely on existing manual lexical resources such as WordNet. To address these challenges, we propose a method to combine distributed representations of words and sentences into summarization task.

Single document summarization using word and sentence embeddings