Home/ Questions/Q 4251126
In Process

The Archive Base Latest Questions

Editorial Team
  • 0
Asked: 2026-05-21T04:37:11+00:00

I am planning to attend a project oriented advanced summer workshop here in India

  • 0

I am planning to attend a project oriented advanced summer workshop here in India on Natural Language Processing.
Before start of the workshop, I have to make a project preference out of the following four areas about which I have limited knowledge.

Machine Translation Develop an English-Indian language translation
system.

Parsing Build an Indian Language (IL) Parser.

Morphological Analysis Develop and test Morphological Analyzers for
Indian Languages.

Speech Spoken Dialog Systems, Emotion/Prosody Detection, Synthesis
and Conversion

I have taken a course in Artificial Intelligence where NLP was introduced and fundamental sub-topics like POS tagging(Transformation Based Learning), word prediction using N-grams, Hidden Markov Models, Viterbi Algorithm, Natural Language Parsing, Context Free Grammar, CKY Algorithm were covered.

I understand this is a slightly vague question and the choice would depend primarily on my interests, but would appreciate guidance on which area would be better in terms of the research scope, practical application, industry opportunities etc.

EDIT: Application of skills/experience acquired while working on the project, outside NLP would also be a factor in the decision.

Share

Leave an answer

You must login to add an answer.

Need An Account,

1 Answer

  1. Editorial Team

    Let’s first group the four options as first three in one category-NL-Text and another fourth in other-NL-Speech, as the skill set and inclination needed to pursue those are slightly different.
    So first use the criteria of do you like working in the first group or second. Once you are done with that and you choose Speech. you are done.
    But if you are on other side, now there are two categories MT and remaining MA and IL Parsing. Building a MT will focus on using the ready made components and adapt them to you language pair. If you are supposed to do it statistically, it’s little more on data side and you don’t gain much, I mean learn/work much as in other two.If that is rule based, there are very few making it big- but you will learn a lot. Building an IL parser is a good option and might be useful for future task where will have lot of data in IL and then text processing industry will flourish. So considering future scope in industry consider my +1. Same case is with Morphological Analysis.

    • 0