Popular GitHub Repositories for Stackoverflow

【不定期更新】收集整理的一些网站中（如知乎、Quora、Reddit、Stack Exchange等）与深度学习、机器学习、强化学习、数据科学相关的有价值的问题

FlexiGPT plugin for VSCode. Interact with AI models as a power user

A Pre-trained BERT on StackOverflow Corpus

FlexiGPT plugin for VSCode. Interact with AI models as a power user

Word2Vec encodings based search engine for Stackoverflow questions

Recommendation system for Stack Overflow unanswered questions

Argument Mining from online debates and hybrid-semantics reasoning

Predicting tags for StackOverflow Questions

Pretrained word embeddings on a corpus of Android app code

A Django Based Web App for predicting a given query belongs to which language i.e 'R', 'java', 'javascript', 'PHP', 'python'. The model is trained using LSTM's with a training accuracy of 97% and testing accuracy of 80%. The data that the model is trained with queries collected from a dataset in Kaggle originally extracted from StackOverflow.

A virtual voice assistant using Speech Recognition and Google API's that can send, receive e-mails, Whatsapp messages to multiple people, setting up Google calendar event, surfing Youtube , Google, Stackoverflow, playing music, etc.

Word2Vec on Tags of Stackoverflow.

一个基于Java的模型上下文协议（MCP）中介器实现，它能够在MCP客户端和服务器之间实现无缝集成。

Find duplicate questions on StackOverflow by their embeddings. From the Natural Language Processing course - Coursera's Advanced Machine Learning specialization.

Using word embedding to find duplicate questions from StackOverflow. Week 3 course of Natural Language Processing course from Coursera.

An AI-powered assistant that searches StackOverflow and provides intelligent summaries of programming solutions using Google's Gemini AI.

We classified Stack Overflow Python questions from 2008-2016 with Natural Language Processing and Deep Learning. Using Regular Expressions, we removed HTML tags and punctuation. We also utilized spaCy to tokenize, lemmatize and remove stop words. Using Keras, we built a 4 layered artificial neural network with a 20% dropout rate using relu and softmax activation functions. We also utilized the adam optimizer and categorical cross-entropy loss function which classified 11 tags 88% successfully.

Contrastive Learning for API Aspect Analysis

Comparaison de différentes approches pour la conception d'un outil de suggestion de mots-clés pour les messages rédigés sur un forum. Travail réalisé dans un cadre MLOps.