mathematics Article Lexical Sense Labeling and Sentiment Potential Analysis Using Corpus-Based Dependency Graph Tajana Ban Kirigin 1, * , Sanda Bujaˇ ci´ c Babi´ c 1 and Benedikt Perak 2   Citation: Ban Kirigin, T.; Bujaˇ ci´ c Babi´ c, S.; Perak, B. Lexical Sense Labeling and Sentiment Potential Analysis Using Corpus-Based Dependency Graph. Mathematics 2021, 9, 1449. https:// doi.org/10.3390/math9121449 Academic Editors: Jonatan Lerga, Ljubisa Stankovic, Nicoletta Saulig and Cornel Ioana Received: 20 April 2021 Accepted: 12 June 2021 Published: 21 June 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affil- iations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). 1 Department of Mathematics, University of Rijeka, R. Matejˇ ci´ c 2, 51000 Rijeka, Croatia; sbujacic@uniri.hr 2 Faculty of Humanities and Social Sciences, University of Rijeka, Sveuˇ cilišna Avenija 4, 51000 Rijeka, Croatia; bperak@uniri.hr * Correspondence: bank@uniri.hr Abstract: This paper describes a graph method for labeling word senses and identifying lexical sentiment potential by integrating the corpus-based syntactic-semantic dependency graph layer, lexical semantic and sentiment dictionaries. The method, implemented as ConGraCNet application on different languages and corpora, projects a semantic function onto a particular syntactical de- pendency layer and constructs a seed lexeme graph with collocates of high conceptual similarity. The seed lexeme graph is clustered into subgraphs that reveal the polysemous semantic nature of a lexeme in a corpus. The construction of the WordNet hypernym graph provides a set of synset labels that generalize the senses for each lexical cluster. By integrating sentiment dictionaries, we introduce graph propagation methods for sentiment analysis. Original dictionary sentiment values are integrated into ConGraCNet lexical graph to compute sentiment values of node lexemes and lexical clusters, and identify the sentiment potential of lexemes with respect to a corpus. The method can be used to resolve sparseness of sentiment dictionaries and enrich the sentiment evaluation of lexical structures in sentiment dictionaries by revealing the relative sentiment potential of polysemous lexemes with respect to a specific corpus. The proposed approach has the potential to be used as a complementary method to other NLP resources and tasks, including word disambiguation, domain relatedness, sense structure, metaphoricity, as well as a cross- and intra-cultural discourse variations of prototypical conceptualization patterns and knowledge representations. Keywords: lexical graph analysis; corpus; knowledge representation and reasoning; affective com- puting; sentiment analysis 1. Introduction The expression of feelings and moods in language is one of the foundations of social communication and interaction of personal and cultural values. Linguistic expressions activate feeling as an emergent cognitive interpretation of the components of the utterance: words and their syntactic organization. According to research in cognitive science [1] and linguistics [2], the process of affective evaluation of a symbolic code is an important component in emergent complex phenomena of creating a sense. Without recognizing, integrating and appraising an affective value in the linguistically articulated conceptual content, be it a rough grained positive vs. negative classification or a nuanced emotional categorization, there is no real comprehension of the text. For humans, the process of expe- riencing affective quality is evolutionary hardwired, sub-conscious trait that is activated by social interaction. However, humans also have difficulties objectively assessing the affective value of an utterance. On the other hand, for a sequence matching quantitative system, a computer, this is an even more difficult task. Nonetheless, in recent years there has been a surge of natural language processing (NLP) techniques and resources that address affective and subjective phenomena in text analysis. Although reductive, the resources are becoming more extensive and versatile due to their quantitative nature. At the same time, graph theory, as a branch of discrete Mathematics 2021, 9, 1449. https://doi.org/10.3390/math9121449 https://www.mdpi.com/journal/mathematics