Probability Distributions.- Linear Models for Regression.- Linear Models for Classification.- Neural Networks.- Kernel Methods.- Sparse Kernel Machines.- Graphical Models.- Mixture Models and EM.- Approximate Inference.- Sampling Methods.- Continuous Latent Variables.- Sequential Data.- Combining Models.

Pattern Recognition and Machine Learning

Soil Chemical Analysis

Statistical Methods For Agricultural Workers

Principles of Plant Breeding

We present the results of the WMT13 shared tasks, which included a translation task, a task for run-time estimation of machine translation quality, and an unofficial metrics task. This year, 143 machine translation systems were submitted to the ten translation tasks from 23 institutions. An additional 6 anonymized systems were included, and were then evaluated both automatically and manually, in our largest manual evaluation to date. The quality estimation task had four subtasks, with a total of 14 teams, submitting 55 entries.

/pdf/findings-of-the-2013-workshop-on-statistical-machine-4e7lke5l51.pdf

Findings of the 2013 Workshop on Statistical Machine Translation

Transliteration is the process of transcribing words from a source script to a target script. These words can be content words or proper nouns. They may be of local or foreign origin. In this paper we present a more discerning method which applies different techniques based on the word origin. The techniques used also take into account the properties of the scripts. Our approach does not require training data on the target side, while it uses more sophisticated techniques on the source side. Fuzzy string matching is used to compensate for lack of training on the target side. We have evaluated on two Indian languages and have achieved substantially better results (increase of up to 0.44 in MRR) than the baseline and comparable to the state of the art. Our experiments clearly show that word origin is an important factor in achieving higher accuracy in transliteration.

/pdf/a-more-discerning-and-adaptable-multilingual-transliteration-3v2ifk8qtc.pdf

A More Discerning and Adaptable Multilingual Transliteration Mechanism for Indian Languages.

In this paper we first present a brief discussion of the problem of Named Entity Recognition (NER) in the context of the IJCNLP workshop on NER for South and South East Asian (SSEA) languages 1 . We also present a short report on the development of a named entity annotated corpus in five South Asian language, namely Hindi, Bengali, Telugu, Oriya and Urdu. We present some details about a new named entity tagset used for this corpus and describe the annotation guidelines. Since the corpus was used for a shared task, we also explain the evaluation measures used for the task. We then present the results of our experiments on a baseline which uses a maximum entropy based approach. Finally, we give an overview of the papers to be presented at the workshop, including those from the shared task track. We discuss the results obtained by teams participating in the task and compare their results with the baseline results.

/pdf/named-entity-recognition-for-south-and-south-east-asian-4u3vagsg75.pdf

Named Entity Recognition for South and South East Asian Languages: Taking Stock

Letter-to-phoneme conversion plays an important role in several applications. It can be a difficult task because the mapping from letters to phonemes can be many-to-many. We present a language independent letter-to-phoneme conversion approach which is based on the popular phrase based Statistical Machine Translation techniques. The results of our experiments clearly demonstrate that such techniques can be used effectively for letter-to-phoneme conversion. Our results show an overall improvement of 5.8% over the baseline and are comparable to the state of the art. We also propose a measure to estimate the difficulty level of L2P task for a language.

/pdf/modeling-letter-to-phoneme-conversion-as-a-phrase-based-25yp3sd67c.pdf

Modeling Letter-to-Phoneme Conversion as a Phrase Based Statistical Machine Translation Problem with Minimum Error Rate Training

To determine how close two language models (e.g., n-grams models) are, we can use several distance measures. If we can represent the models as distributions, then the similarity is basically the similarity of distributions. And a number of measures are based on information theoretic approach. In this paper we present some experiments on using such similarity measures for an old Natural Language Processing (NLP) problem. One of the measures considered is perhaps a novel one, which we have called mutual cross entropy. Other measures are either well known or based on well known measures, but the results obtained with them vis-avis one-another might help in gaining an insight into how similarity measures work in practice.

The first step in processing a text is to identify the language and encoding of its contents. This is a practical problem since for many languages, there are no universally followed text encoding standards. The method we have used in this paper for language and encoding identification uses pruned character n-grams, alone as well augmented with word n-grams. This method seems to give results comparable to other methods.

/pdf/study-of-some-distance-measures-for-language-and-encoding-2nqsi4g307.pdf

Study of Some Distance Measures for Language and Encoding Identification

Text on the Web is available in numerous languages and encodings, often not according to any standards. The number of multilingual documents on the Web is also increasing. The problem of identifying the languages and encodings in a multilingual document and marking portions of a document with them has not been addressed so far. We present an exploration of this problem, the implied or required assumptions, and a solution. The problem can be divided into three parts: monolingual identification, enumeration of languages and identification of the language of every portion. For enumeration, we have been able to get a precision of 96.20%. We also experimented on language identification of each word. Given correct enumeration, we could obtain type precision of 90.91% and token precision of 86.80%. Finally, we show how precision is affected by language distance.

/pdf/identification-of-languages-and-encodings-in-a-multilingual-4wj5nkhey1.pdf

Anil Kumar Singh

Papers

A More Discerning and Adaptable Multilingual Transliteration Mechanism for Indian Languages.

Named Entity Recognition for South and South East Asian Languages: Taking Stock

Modeling Letter-to-Phoneme Conversion as a Phrase Based Statistical Machine Translation Problem with Minimum Error Rate Training

Study of Some Distance Measures for Language and Encoding Identification

Identification of Languages and Encodings in a Multilingual Document