[go: up one dir, main page]

Skip to content
View anjali-rgpt's full-sized avatar
  • Stanford University

Block or report anjali-rgpt

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
anjali-rgpt/README.md

Hi there 👋

Specialization : Natural Language Processing, Prompt Engineering, Natural Language Understanding, Human-Centered Artificial Intelligence Design, Ethics in NLP

Preferred Programming Languages : Python 3, Java, C

Domains of Active Exploration : Computational Linguistics, Educational Technology, Computational Social Science

Some key projects from the past include :

Topic Classification for Garage Reviews - An ML Task [^1] A project developed for automatic tagging of customer reviews using unsupervised topic modelling (LDA, GSDMM, CTM) and supervised multi-class, multi-label classification (Naive Bayes, Logistic Regression, SVM).

Spotify Similarity-Based Recommender System [^2] Project for machine learning and data mining course exploring the Spotify Dataset, comparing clustering methods, and determining the best similarity measure to be used for recommendation of new songs.

Grouping and recommending literature based on their content [^3] An NLP project utilising LDA to group books into themes based on their topics and to recommend similar books to the user. Web scraping from Wikipedia is used to get the plot of the book provided in test data.

RESEARCH : Efficient Autocomplete Algorithms [^4] An exploratory NLP project looking at different solutions for autocomplete, including FastText embeddings and non-linear data structures, to increase the use of sustainable and energy-efficient machine learning in code without compromising on result quality.

RESEARCH: Automatic Evaluation of Short-Answer Responses using Clustering and Summarization [^5]Snippets of Jupyter notebooks exploring clustering and summarization methods on examinations from students in introductory programmatic reasoning classes.

RESEARCH: Psycholinguistics Experiment as Part of LLM Idiom Interpretation Study [^6]Complete project repository containing experiment setup, data, and analysis scripts for psycholinguistic study on unfamiliar idiom interpretation in human participants.

Pipeline : A preprocessing pipeline for general-purpose tokenization, stopword removal, POS-tagging, and n-gram generation [^6] A useful piece of code with multiple functions that can be conveiently accessed using a pipeline method. Work in Progress to convert it into an encapsulated version of itself.

Pinned Loading

  1. Autocomplete Autocomplete Public

    A repository containing the code that was developed as part of an exploratory NLP project with 24/7.ai, based on completing words and predicting the next words.

    Jupyter Notebook

  2. NLP-Projects NLP-Projects Public

    Projects on text clustering / summarization, topic classification (topic modeling) for literature recommendations and review tagging

    Jupyter Notebook

  3. MediaSum-StackExchange-Dialogue-Analysis MediaSum-StackExchange-Dialogue-Analysis Public

    Using MediaSum Transcripts, analyse them and find out the primary topic groups to which the questions belong. Generate word clouds for the same. From StackExchange Data Dump as the data store, loca…

    Jupyter Notebook

  4. a98zhang/argue-better a98zhang/argue-better Public

    Argue Better: Using Language Models to Identify Weak Argument Structure in Essays and Provide Examples for Correction

    Jupyter Notebook 1

  5. CS224C-project CS224C-project Public

    Project designed to assess if GPT-3 can be a good course forum assistant. We compare human teaching team responses to common course forum questions for old Stanford Online courses to zero-shot GPT-…

    Jupyter Notebook