Processing of natural language for extraction of the meaning is a challenge in the field of artificial intelligence. Research work in this area is being carried out in most of the Indian and foreign languages by analyzing the grammatical aspect of these languages. Sanskrit, a language that possesses a definite rule-based structure given by Panini, has a great potential in the field of semantic extraction. Hence, Sanskrit and computational linguistic are strongly associated. As given in the grammar of Sanskrit language, its case endings are strong identifiers of the respective word in the sentence. To extract the semantic from the language, dependencies amongst the words of a sentence are developed, and the semantic role of words is identified (e.g., agent, object, etc.). In this work, an algorithm has been developed for creating a dependency-based structure for the sentence in Sanskrit by analyzing the features given by the Part of Speech (POS) tags. Dependency Tags (DTs) are used to relate the verb with other words in a sentence. POS tags give the syntactic information and DT gives the semantic information. Mapping between the two is established in the proposed algorithm and its analysis is done. Sanskrit, being an order-free language, imposes a great challenge for the development of dependency-based structure for the sentence. This paper is an effort in the same direction. Such representations are useful for applications such as knowledge representation, query-based system and machine learning.

Dependency Grammar (DG) plays an important role in theoretical linguistics and natural language processing. There has been an increasing interest in dependency-based representation in natural language processing as it aims towards extracting the semantic from the language. In this paper, we have presented a state-of-the-art dependency-based processing of Sanskrit language. Although this concept has been rediscovered, it has its links to Panini’s grammar of Sanskrit which was presented several centuries before the common era (Kruijff, 2002) and to the theories of grammar (Covington, 2001). Dependency grammar has largely developed as a form of syntactic representation used by traditional grammarians, especially in Europe, in classical and Slavic domains (Joakim, 2005). Dependency concepts are found in traditional Latin, Arabic and Sanskrit grammar, among others, which have attracted computer researchers for the past four decades, but there has been little systematic study of dependency parsing (Covington, 2001).

Parsers of languages are of two types—constituency parser based on phrasestructure grammar and dependency parser based on DG. In phrase-structure grammar, theory of automata is used, whereas dependency grammar establishes a word-to-word link between the words of a sentence. Works of some researchers (Covington, 2001; and Hellwig, 2007) have differentiated both types of parsers and also developed constraintbased equivalence relation between them, which is popularly known as X-bar theory (Covington, 2001). Dependency parser is best suited for order-free languages like Sanskrit and Hindi, hence we have taken dependency-based analysis for the text. In this paper, we present an algorithm for creating the dependency structure for a POS-tagged Sanskrit sentence. The common question asked is why Sanskrit is used as an input language. In India, most of the languages, especially Hindi, also draw their features from Sanskrit. With its rich morphological structure and efficient grammar , this language has a great potential in the field of computational linguistic, especially in applications where semantic analysis, knowledge representation and machine learning are the key issues (Akshar and Rajeev, 1993; Akshar et al., 1995; and Narsingh, 2005). It has been stated as the language with the most systematic grammar suitable for knowledge representation (Briggs, 1995).

Computer Sciences Journal, Semantic Analysis, Dependency Tree Construction, Sanskrit Language, Artificial Intelligence, Natural Language Processing, Indian Languages, Dependency Structures, Dependency Tag, Morphological Structure, Theoretical Linguistics.