Author: abayramova99

There are different aspects of research process. To find and use the right dataset is also important, especially for the projects related to Natural Language Processing. In most of the NLP projects it is crucial to use real dataset to get effective implementation and better outcomes.

Developing Dialog Management Systems is one of the such projects. To build conversational agent which is capable of handling complex dialog turns and continue human-like conversation, dataset used to train the models also should contain some conversational patterns. For example, text of messages, dataset from the messengers such as WhatsApp, Facebook or online customer service are good examples for possible datasets. Additionally, dataset should reflect the characteristics of the project domain. For example, chatbots of the touristic companies must be able to handle common customer requests, related to the touristic places, etc.

In my work, I am planning to use real human to human messages in Azerbaijani language. It is noisy dataset which contains misspellings, noise of internet data and incomplete sentences. Additionally, agglutinative nature of Azerbaijani language, for instance, having several morphological forms of the same words, should also be considered. On the other hand, this dataset has applicable for chatbots structure, in the form of questions and answers.

There are many aspects should be taken into account while choosing the research topic for MS Project. First of all, it must be achievable in the frame of time and resources of the student and the program. Secondly, it must be novel topic in which society (or some part of it) will be interested in. From my perspective, it means that either this project will be useful for the people in their lives or some industrial/scientific organization will benefit from it.

For myself, with outcomes of the research topic I should accomplish my academic goals and personal interests. I would like to realize my ideas in the field of Computer Science and develop a software which will be implemented in real cases. The ideal topic, in my opinion, is the one which combines together my area of interest and demand of the industry, especially in my home country. Additionally, it must meet requirements listed above.

While working on the research project several questions should be answered. We should define the problem itself, if it is valuable for society, etc. On the other hand, we should think about approaches which are already applied and novel ones, how to get qualitative dataset to carry out experiments, what will be risk of those experiments and whole project. Last but not least, we should think about the what the result would be and how to evaluate its success.

I am Aygul Bayramova, graduate student in Computer Science at The George Washington University and data scientist at Unibank CB, Azerbaijan.

I was born and spent most of my childhood in Gusar, Azerbaijan. It is a small town with wonderful nature in nothern part of Azerbaijan, near Caucasian Mountains.

Within Computer Science I am interested in Machine Learning, Natural Language Processing, Robotics and Game Theory . Outside of Computer Science, I like playing chess and win 😁. Additionally, I like to listen to music and spend time with my friends.

Welcome to your brand new blog at GW Blogs.

To get started, simply log in, edit or delete this post and check out all the other options available to you.

For assistance, visit our comprehensive support site, check out our Edublogs User Guide guide or stop by The Edublogs Forums to chat with other edubloggers.

You can also subscribe to our brilliant free publication, The Edublogger, which is jammed with helpful tips, ideas and more.