Examination of the Corpus of Attitude Markers in the Persian Language

Attitude markers[1] are a set of expressions in a language that speakers use to clarify and explain their feelings and viewpoints. They primarily appear as adverbs, sentence-level adverbs, adjectives, verbs, and nouns. Words or phrases such as: unfortunately, clearly, certainly, surprisingly, easily, unconsciously, by force, I believe, etc., are examples of attitude markers that express the speaker's feelings and attitudes. In fact, attitude markers reinforce the intended meaning of the speaker and provide information about the discourse structure of the text, facilitating the connection between the components of the text and playing a crucial role in interpreting and understanding the meaning of the text. These elements are not merely used for text coherence. For example, the conjunction "and" in the sentence "Mary went to Germany and lived there" only connects two sentences, whereas the word "madly" in the sentence "He drove madly on the two-way road" expresses the speaker's feeling about the action performed and is considered a type of attitude marker. The aim of the present research is to identify and classify attitude markers in the Persian language. For this purpose, a list of attitude markers will be extracted from relevant sources (by relevant sources, we mean specialized books and articles in Persian and English authored by linguists who have studied attitude markers) and simultaneously, the classification of attitude markers will be conducted, placing them into various categories. After identifying and classifying attitude markers in Persian, the frequency of these markers will be extracted from both general and specialized corpora to provide an overview of the usage of each attitude marker in both corpora. For example, we will ultimately arrive at information such as: the frequency of the adverb "certainly," which indicates certainty, being 20% in the general corpus and 30% in the specialized corpus. Such information is useful for research that examines the language of science and shows how much each of the attitude markers is used by authors of specialized and general texts. The author emphasizes that the extraction of the frequency of attitude markers from the general and specialized corpora will be conducted after the extraction and classification of the attitude markers and is not the main objective of the research. This research can primarily be used in discourse tagging of corpus data (assigning labels such as affirmation, acceptance, condition, certainty, cause, result, opposition, judgment, praise, etc.), information retrieval from texts, automated question-and-answer systems based on viewpoints, extracting evidence for a claim from relevant texts, and so on. Specifically, the results of this research can be used for tagging data in the existing corpora of the IranDoc system. Another very important application of this research is in tools and linguistic processes related to sentiment analysis[2].

This research will be conducted in two steps:

Step One: Extracting attitude markers from relevant sources and classifying them.

Step Two: Extracting the frequency of attitude markers from one specialized corpus and one general corpus.