preview

Lexicon Generation Research Paper

Decent Essays

2 Related Work Several well-regarded manually compiled sentiment lexicons do exist [2-4]. Due to their high cost in terms of time and effort, however, a large volume of research concentrated on automated sentiment lexicon generation has emerged in the past few years. Automated sentiment lexicon generation branches in two main directions: dictionarybased and corpus-based. The volume editors, usually the program chairs, will be your main points of contact for the preparation of the volume. The dictionary-based approach to generate a sentiment lexicon involves leveraging linguistic resources and online dictionaries (e.g. WordNet) to automatically tag words with their corresponding semantic orientations [5, 6]. This generates a general, …show more content…

Wan [12] also attempts a similar technique to derive a Chinese sentiment lexicon from bilingual resources, with relatively poor accuracy. Saif et al. [13] and Salameh et al. [14] investigate this issue in detail and mention that, when text is translated from a source language to a target language, the sentiment of terms is preserved to varying degrees, with great reliance on the machine translation technique involved. Tan et al. [15] compile a Malay sentiment lexicon by manually translating terms in the Affin lexicon [16] to their Malay counterparts, and supplement the lexicon with slang terms commonly used in Malay social media posts. Shamsudin et al. [17] manually compile a sentiment lexicon using WordNet Bahasa to classify Malay social media posts. This defeats the purpose for an automated means of lexicon construction. Malay sentiment analysis has started to witness rapid progress both in industry (e.g. [18]) 2016 and academia (e.g. [19]) during the past few years. [20] develop a knowledge base approach combined with supervised classifiers for sentiment classification of Malay text. [21] first construct a lexicon for a particular Malaysian dialect, i.e., the Sabah language., and employ it to categorize a social media posts dataset. [22] classify Malay news headlines using a series of supervised classifiers. [23] investigate

Get Access