Dış Kaynaklı Proje Süreçleri Yönetim Sistemi
Samet R. (Yürütücü), Emrah Ş. , Özkan M.
Advancements in science and technology have rapidly propelled the field of text processing for a diverse range of languages. For widely resourced languages, such as English, Turkish, and others, there exist robust corpora, specialized dictionaries for textual analysis, and extensive datasets to support sentiment polarity detection. These resources facilitate efficient language processing and enable accurate sentiment analysis, allowing for the identification of nuanced emotional tones within a text. However, for Azerbaijani such resources and analytical frameworks are either scarce or non-existent. This lack of foundational frameworks and resources for Azerbaijani presents significant challenges in analyzing sentiment accurately and comprehensively, as cultural and linguistic nuances specific to this language remain underrepresented in available technologies.
The proposed project is designed to address the challenges encountered when conducting sentiment analysis for Azerbaijani. Sentiment analysis presents a significant challenge, particularly in accurately identifying the positive, negative, or neutral connotations of different words and determining the emotional tendency within a text based on these meanings. With the rapid increase in available online content, such analysis not only involves processing large amounts of text from various data sources like social media or news content but also requires implementing comprehensive analytical frameworks that understand the cultural and linguistic nuances of the language.
This project aims to fill a critical gap, as Azerbaijani is a language with limited linguistic resources available for natural language processing and sentiment analysis. The scarcity of analytical frameworks makes it difficult to conduct accurate and comprehensive sentiment analysis, thus highlighting the urgent need for the development of a dedicated polarity lexicon. To address this need, the project aims to develop an Azerbaijani Polarity Lexicon and create a sentiment analysis framework. This framework will be made accessible through an online platform that will enable users to analyze texts and determine their emotional tendencies, making it particularly valuable for social media analysis, content monitoring, and market research. The project team will compile a comprehensive dataset of Azerbaijani words labeled with their positive, negative, or neutral sentiments, which will serve as the foundation for the sentiment analysis system. This linguistic resource will be used to create effective natural language processing solutions for analyzing sentiment in Azerbaijani text.
The project currently resides at Technology Readiness Level (TRL) 2, characterized by the formulation of a technology concept. Ongoing efforts are focused on concept development, including the creation of a specialized lexicon and analytical framework. Subsequent project stages will involve the development and testing of the framework and its implementation as a web-based application, with the aim of advancing to higher TRLs. The project aspires to achieve at least TRL 5, signifying technology validation within an industrially relevant environment, a key milestone for enabling technologies.