Sentence examples for text similarity from inspiring English sources

Of the 34 highly similar pairs (full text similarity ratio >0.85) in PMC that we examined, none would be considered "unethical" by the average scientist because they were updates or multi-part publications, etc.

In Figure 2, for the "dissimilar abstract" group, the frequency of citation pairs drops sharply as the full text similarity increases from 0.4 to 0.55, whereas for the "similar abstract" group, the frequency of citation pairs peaks when the full text similarity is close to 0.55.

We built clustered similar phenotypes using the vector space model for text similarity and showed that the emerging clusters are biologically coherent, enabling gene function prediction with a precision of over 70%.

For these cases, the use of a text similarity tool (using a bag of word approach like eTBLAST or short similar sentences with SIPs) would fail because text similarity does not account for natural syntactic inconsistencies such as synonym use or grammatical variations.

This strategy is very useful to analyze text similarity in huge collections ([Rajaraman and Ullman 2012]).

We use text similarity matching techniques applied on stations' names to double check our results.

In our previous study, the text similarity-based information retrieval search engine eTBLAST [1] was tuned with the MEDLINE abstract dataset [2] to create Déjà vu, a publicly available database of over 70,000 highly similar biomedical citations [3].

One test of the interpretation of the semantic factors was obtained by using LSA (http://lsa.colorado.edu/), which applies singular value decomposition to corpus-based metrics to provide a high-dimensional (300 in our case) representation of inter-text similarity [26].

The use of two text-similarity-recognition programs also improved the rate of detection and, in some theses, significantly increased the classification of the gravity of the plagiarism encountered.

Of the theses, 138 were inspected by both TurnItIn and Urkund, two highly rated text-similarity-recognition (TSR) programs (Weber-Wulff et al. 2013, p. 9), whereas the 12 M.A. end-of-course monographs from UEM's economics department were only inspected by Urkund.

To evaluate the correctness of incremental representation, social text similarity/relatedness, linguistic tasks, network event detection, social user multi-label classification and user clustering for social network analysis are employed as benchmarks in this paper.