用于防暴游戏的社交媒体分析，在软件补丁之前和之后分析玩家情绪，以优化游戏更新，提高玩家保留率和满意度。利用API抓取和LDA主题建模从多个子红字中进行数据提取和情感分析。
social media analytics for Riot Gaming, analyzing player sentiment pre and post software patches to optimize game updates, enhancing player retention and satisfaction. Utilized API scraping and LDA Topic Modeling for data extraction and sentiment analysis from multiple subreddits. (2024-04-04, Jupyter Notebook, 0KB, 下载0次)
分析并将BBC 2004-2005年的新闻文章分为五个主题类别：商业、娱乐、政治、体育和科技使用...，
Analyze and cluster BBC news articles from 2004-2005 into five topical categories: business, entertainment, politics, sport, and tech using MiniBatchKMeans and TF-IDF features. (2023-10-10, Jupyter Notebook, 0KB, 下载0次)
玩着编写标记器，
playing with writing a tokenizer, (2023-09-20, Jupyter Notebook, 0KB, 下载0次)
使用Riot Api和仙后座包来争论NA、EUW和KR Ladder上挑战者玩家的数据。
Using Riot Api and Cassiopeia Package to wrangle data on Challenger players on the NA, EUW, and KR Ladder. (2021-12-31, Jupyter Notebook, 574KB, 下载0次)
斯坦福大学自然语言处理与深度学习课程cs224n作业2的编码部分
Coding part of cs224n Assignment 2, Stanford s Natural Language Processing with Deep Learning Course (2019-10-09, Jupyter Notebook, 344KB, 下载0次)
玩Word2Verc和t-SNE
Playing around with Word2Vec and t-SNE (2016-02-28, Jupyter Notebook, 1430KB, 下载0次)
美国有线电视新闻网使用Glove和Word2Vec对斯坦福情绪树库数据集进行情绪分析
CNN for sentiment analysis over the Stanford Sentiment Treebank dataset, using Glove and Word2Vec (2018-03-08, Jupyter Notebook, 841KB, 下载0次)
...类别，即商业、娱乐、政治、体育和科技。使用这种方法，我们能够找到...
Article classification and summarization have a genuinely backhanded connection as article classification fall into classification issues rather than summarization, where it is treated as an issue of semantics. A significant piece of the summarization procedure is the recognition of the point or subjects that are examined in an irregular (2020-08-04, Jupyter Notebook, 5247KB, 下载0次)
玩古腾堡计划的短信。
Playing around with texts from Project Gutenberg. (2018-03-15, Jupyter Notebook, 4605KB, 下载0次)
我们提出并探索了一个基于维基百科数据和斯坦福问答数据集的QA系统。我们展示...
We propose and explore a QA system based on data form Wikipedia and the Stanford Question Answering Dataset. We show different approaches to document retrival as well as reading. Including tf-idf and term frequency for retrival. Language models such as word2vec/glove, infersent(sentence2vec by fb research),... Furthermore to we apply different (2020-01-18, Jupyter Notebook, 82417KB, 下载0次)