Resources for Turkish natural language processing: A critical survey.

Area of Science:

• Computational Linguistics
• Natural Language Processing
• Lexicography

Background:

• The Turkish language has a growing need for robust digital linguistic resources.
• Existing corpora and lexical data for Turkish are fragmented and not always accessible.
• Advancements in Natural Language Processing (NLP) necessitate comprehensive resource availability.

Purpose of the Study:

• To conduct a thorough survey of existing Turkish corpora and lexical resources.
• To focus on resources that are publicly accessible for broader research use.
• To provide recommendations and identify critical data gaps for future development.

Main Methods:

• Systematic literature review of linguistic resource databases.
• Analysis of publicly available online repositories for Turkish language data.
• Categorization and evaluation of identified corpora and lexical datasets.

Main Results:

• A comprehensive overview of currently available Turkish corpora and lexical resources.
• Identification of specific areas with insufficient data for NLP tasks.
• Documentation of resources suitable for immediate use in research and application development.

Conclusions:

• There is a significant need for curated and expanded Turkish language resources.
• Addressing identified data gaps is crucial for advancing Turkish NLP.
• Recommendations are provided to guide future resource creation and data collection efforts.