Akshar Bharati

Parsing Free Word Order Languages in the Paninian Framework

There is a need to develop a suitable computational grammar formalism for free word order languages for two reasons: First, a suitably designed formalism is likely to be more efficient. Second, such a formalism is also likely to be linguistically more elegant and satisfying. In this paper, we describe such a formalism, called the Paninian framework, that has been successfully applied to Indian languages.This paper shows that the Paninian framework applied to modern Indian languages gives an elegant account of the relation between surface form (vibhakti) and semantic (karaka) roles. The mapping is elegant and compact. The same basic account also explains active-passives and complex sentences. This suggests that the solution is not just adhoc but has a deeper underlying unity.A constraint based parser is described for the framework. The constraints problem reduces to bipartite graph matching problem because of the nature of constraints. Efficient solutions are known for these problems.It is interesting to observe that such a parser (designed for free word order languages) compares well in asymptotic time complexity with the parser for context free grammars (CFGs) which are basically designed for positional languages.

Two stage constraint based hybrid approach to free word order language dependency parsing

The paper describes the overall design of a new two stage constraint based hybrid approach to dependency parsing. We define the two stages and show how different grammatical construct are parsed at appropriate stages. This division leads to selective identification and resolution of specific dependency relations at the two stages. Furthermore, we show how the use of hard constraints and soft constraints helps us build an efficient and robust hybrid parser. Finally, we evaluate the implemented parser on Hindi and compare the results with that of two data driven dependency parsers.

Insights into Non-projectivity in Hindi

Large scale efforts are underway to create dependency treebanks and parsers for Hindi and other Indian languages. Hindi, being a morphologically rich, flexible word order language, brings challenges such as handling non-projectivity in parsing. In this work, we look at non-projectivity in Hyderabad Dependency Treebank (HyDT) for Hindi. Non-projectivity has been analysed from two perspectives: graph properties that restrict non-projectivity and linguistic phenomenon behind non-projectivity in HyDT. Since Hindi has ample instances of non-projectivity (14% of all structures in HyDT are non-projective), it presents a case for an in depth study of this phenomenon for a better insight, from both of these perspectives. We have looked at graph constriants like planarity, gap degree, edge degree and well-nestedness on structures in HyDT. We also analyse non-projectivity in Hindi in terms of various linguistic parameters such as the causes of non-projectivity, its rigidity (possibility of reordering) and whether the reordered construction is the natural one.

Simple Parser for Indian Languages in a Dependency Framework

This paper is an attempt to show that an intermediary level of analysis is an effective way for carrying out various NLP tasks for linguistically similar languages. We describe a process for developing a simple parser for doing such tasks. This parser uses a grammar driven approach to annotate dependency relations (both inter and intra chunk) at an intermediary level. Ease in identifying a particular dependency relation dictates the degree of analysis reached by the parser. To establish efficiency of the simple parser we show the improvement in its results over previous grammar driven dependency parsing approaches for Indian languages like Hindi. We also propose the possibility of usefulness of the simple parser for Indian languages that are similar in nature.

Inferring Semantic Roles Using Sub-Categorization Frames and Maximum Entropy Model

In this paper, we propose an approach for inferring semantic role using sub-categorization frames and maximum entropy model. Our approach aims to use the sub-categorization information of the verb to label the mandatory arguments of the verb in various possible ways. The ambiguity between the assignment of roles to mandatory arguments is resolved using the maximum entropy model. The unlabelled mandatory arguments and the optional arguments are labelled directly using the maximum entropy model such that their labels are not one among the frame elements of the sub-categorization frame used. Maximum entropy model is preferred because of its novel approach of smoothing. Using this approach, we obtained an F-measure of 68.14% on the development set of the data provided for the CONLL-2005 shared task. We show that this approach performs well in comparison to an approach which uses only the maximum entropy model.

Computational linguistics in India: an overview

Computational linguistics activities in India are being carried out at many institutions. The activities are centred around development of machine translation systems and lexical resources.

SSF: A Common Representation Scheme for Language Analysis for Language Technology Infrastructure Development

We describe a representation scheme and an analysis engine using that scheme, both of which have been used to develop infrastructure for HLT. The Shakti Standard Format is a readable and robust representation scheme for analysis frameworks and other purposes. The representation is highly extensible. This representation scheme, based on the blackboard architectural model, allows a very wide variety of linguistic and non-linguistic information to be stored in one place and operated upon by any number of processing modules. We show how it has been successfully used for building machine translation systems for several language pairs using the same architecture. It has also been used for creation of language resources such as treebanks and for different kinds of annotation interfaces. There is even a query language designed for this representation. Easily wrappable into XML, it can be used equally well for distributed computing.

Handling Multi-word Expressions without Explicit Linguistic Rules in an MT System

Translation of Multi-word expressions (MWEs) is one of the most chal- lenging tasks of a Machine translation (MT) system. In this paper, we present an inno- vative technique for dealing with MWEs in the context of MT. The technique permits bilinguals to give translations of MWEs in the form of patterns, without requiring them to be trained linguistically. The interpretation of the pat terns is done by a dynamic ma- chine learning algorithm, which allows the main rule-based MT system to operate based on linguistic rules. Thus, the bilingual patterns (wi thout any explicit linguistic input) are used in conjunction with the main linguistic syst em. This is made possible by the learning pathway templates. These templates need to be specially prepared by trained linguists only once. After that they help to process potentially a large number of patterns. The implemented system is being used with a large-scale rule-based MT system to improve its performance. This framework can also be extended to help example-based or statistical MT systems to deal with MWEs.

AnnCorra: building tree-banks in Indian languages

This paper describes a dependency based tagging scheme for creating tree banks for Indian languages. The scheme has been so designed that it is comprehensive, easy to use with linear notation and economical in typing effort. It is based on Paninian grammatical model.

Natural language processing, complexity theory and logic

In this paper, we have focussed on assignment of meaning to sentences in NL. We have tried to say that theoretical Computer Science in general, and computational complexity theory and logic in particular, have a lot to contribute to NLP. On the other hand NLP has a number of challenging problems which require new ways of thinking and may thus lead to radically new frameworks in theoretical CS, particularly logic. Recent history has several instances where theoretical CS has helped shape research in NLP. At the same time, several open problems in NLP await work, some of which have the potential to throw up radically new frameworks in theoretical Computer Science.