Identifying the most appropriate expansion of acronyms used in wikipedia text

Abstract

SUMMARYDuring the past several years, many studies have been conducted to determine the expansions (the expanded forms or definitions) of acronyms used in specific texts by using linguistic and syntactical approaches to overcome disambiguation problems. An acronym is an abbreviation formed of the initial components of multiple words. Using only these initial parts can result in huge mistakes when a machine conducts experiments to determine an acronym's meaning in given texts. Finding expansions of acronyms is currently not a major concern. However, identifying how a polysemous acronym is precisely being used in a specific text is difficult. To solve disambiguation problem acronyms have, this paper proposes a method to identify the most appropriate expansions of an acronym by analysing co-occurrence words found in articles on Wikipedia. Our goal is not to find an acronym's definition or expansion but to identify the most appropriate expansion of a given acronym as it is used in a given text. Through experiments, we demonstrate the proposed method's efficiency and performance by comparing several results.

Recognizing Indonesian Acronym and Expansion Pairs with Supervised Learning and MapReduce

“…A previously conducted study [18] used SVM to heuristically determine the accurate expansion candidates based on words similar to the acronyms. Moreover, Choi et al [21] introduced a method to extract the acronym and expansion pairs and examined the co-occurrence of words in Wikipedia to solve the problem of polysemous acronyms.…”

“…Many approaches, such as the letter-matching techniques [13], heuristic extraction algorithms [14], hybrid text mining approaches [15], statistical learning approaches using logistic regression [16], supervised learning with weaker constraints [17], machine learning approaches [18], and mining and ranking approaches [19], have been introduced in the previous two decades to identify acronyms and their definitions because the acronym writing and formation rules generally differ for different languages. Furthermore, several methods have been developed to identify accurate acronyms and expansions from specific data sources such as biomedical literature [16], the Internet [20], and Wikipedia [21]. Various approaches to recognize the accurate acronyms and their definitions in specific languages have been introduced in previously conducted studies [22,23].Herein, we extend the studies conducted by Wahyudi et al [23] and Abidin et al [24] and propose the recognition of acronym and expansion pairs from a large Indonesian corpus using machine learning techniques and the Hadoop big data technology.…”