This paper describes the UNIBA team participation in the Cross-Level Semantic Similarity task at SemEval 2014. We propose to combine the output of different semantic similarity measures which exploit Word Sense Disambiguation and Distributional Semantic Models, among other lexical features. The integration of similarity measures is performed by means of two supervised methods based on Gaussian ...

This work uses recursive autoencoders (Socher et al., 2011), word embeddings (Pennington et al., 2014), associative matrices (Schuler, 2014) and lexical overlap features to model human judgments of sentential similarity on SemEval-2015 Task 2: English STS (Agirre et al., 2015). Results show a modest positive correlation between system predictions and human similarity scores, ranking 69th out of...

Requiring only category names as user input is a highly attractive, yet hardly explored, setting for text categorization. Earlier bootstrapping results relied on similarity in LSA space, which captures rather coarse contextual similarity. We suggest improving this scheme by identifying concrete references to the category name’s meaning, obtaining a special variant of lexical expansion.

In this paper we present a novel technique for integrating lexical-semantic knowledge in systems for learning textual entailment recognition rules: the typed anchors. These describe the semantic relations between words across an entailment pair. We integrate our approach in the cross-pair similarity model. Experimental results show that our approach increases performance of cross-pair similarit...

Two types of similarities between words have been studied in the natural language processing community: synonymy and relational similarity. A high degree of similarity exist between synonymous words. On the other hand, a high degree of relational similarity exists between analogous word pairs. We present and empirically test a hypothesis that links these two types of similarities. Specifically,...

With the vision of Semantic Web, the ontology operations such as aligning, merging and mapping have gained much importance. The measuring of similarity between concepts of source ontologies is preprocessing of all these operations. Several techniques have been proposed for measuring similarity between concepts based on their lexical, taxonomic and elementary characteristics but a very little at...

Finding semantic similarity between short biomedical texts, such as article abstracts or experiment descriptions, may provide important information for health researchers. This paper presents a method for calculating text similarity in the biomedical context. The method implements a pairwise concept semantic similarity measure that uses concept definitions and ontology structure. The respective...

In a lexicalized grammar formalism such as Lexicalized Tree-Adjoining Grammar (LTAG), each lexical item is associated with at least one elementary structure (supertag) that localizes syntactic and semantic dependencies. Thus a parser for a lexicalized grammar must search a large set of supertags to choose the right ones to combine for the parse of the sentence. We present techniques for disambi...

The phrase-based Statistical Machine Translation (SMT) approach deals with sentences in isolation, making it difficult to consider discourse context in translation. This poses a challenge for ambiguous words that need discourse knowledge to be correctly translated. We propose a method that benefits from the semantic similarity in lexical chains to improve SMT output by integrating it in a docum...