This is an app that aggregates the most recent and popular tweets, the most interesting videos from youtube, what's happening in the stock and crypto markets, the most recent news, activities and every relevant information happening in realtime on the internet. we also feature a live tv that aggregates videos and livestreams of anime, tv shows, news (again), music, films, and also funny, cool and entertaining cartoons. Trender aggregates data from google, facebook, youtube, twitter, wikipedia, blogs, websites, communities, forums, etc and from this abundance of information and media it builds the ultimate media-internet-entertainment-news network.
Simply put: trender aims to be the frontpage of my internet.
The idea is to build something akin to Trending Topics, but with a much richer and faster and cheaper architecture, that runs fine on localhost without aid of the fatso hadoop.
The trender platform crawls, stores and makes use of search technologies, spiders, databases and artificial inteligence to produce ~accurate, diverse and descriptive information on realtime or at least near to realtime about what's happening in the world. it does so by collecting, processing and aggregating data from the various sources.
Dropwizard rocks
Master hibernate
Build a simple but cool artificial-inteligence to detect trends
Programming & Fun
Put my Solr and Lucene experience to the top
Learn more about search/realtime technologies
How far can my thinkpad go ?
A cool app for myself
Fast, beautifull and interactive apps, featuring videos, pictures, tweets and news frontend consumption, manifold of consumers/producers, crawlers, spiders, ajax, videos, etc
the API allows you to and organize update your data, and querying the system for dynamic information. currently the api is written in java with the dropwizard library and i love it.
search servers, ml & nlp tools. this would be the core of the stuff, in here we want to use SNLP (Sanford Natural Language Processing [Group]) to detect trends using sentiment analysis and some kind of word turnover rate. full text search can be provided via postgres (flexible, fast and convenient) or some other specialized vendor (lucene, elastic, sphinx). the reason we need a full-text search offering it's because it doenst make sense to detect good/bad trends if your searching abilities are poor. so well the search will have some support for Semantic matching, as it's way cool.
that's it
- Natural Language Processing
- Dropwizard Metrics ?
- Trending API
- Crawling API
- Discovery API
- Searching API
- Searching infraestructure optimization (solr)
- Scrapy crawling infraestructure
use spiders to get media content
well, we enjoy the luxury of things happening all the time (crawling activity)
what if we used that to implement some kind of serve->test ( ~live testing ? )
bittorrent architecture & checkout tornet
- list of crawlers
- gigablast
- erlang ebot
- Scrapy
- grequests
- Command-line JSON processor
- awesome-streaming
- jet.hazelcast
- @rwalk straw
- @rwalk blog post on straw
- luwak
- real-time-full-text-search-with-luwak-and-samza
- scalable web crawlers on Apache Storm
- ebot crawler blogpost
- seleniumhq.org
- htaccess snippets
- syntax-highlighter
- twitter-emoji
- yago Knowledge Base
- caffe.berkeleyvision(deep learning framework)
- The Stanford NLP Group
- Stanford CoreNLP
- Learning words (google word2vec)
- google word2vec
- CS 229 Machine Learning
- Exploring LSTMS
- RNN Effectiveness
- IBM Watson
- Facebook ai-negociator-bots
- Quora QMF
- Facebook DeepText
- Facebook ParlAI
- OpenAI
- openai systems
- amazon mechanical turk
- deeplearning.net
- artificial-intelligence-revealed (facebook)
- openweathermap
- thetvdb
- Oil Price
- blockchain
- blockchair
- coinmarketcap
- bravenewcoin
- bitcoin developer guide
- cryptocompare
- fork.lol
- bitcoin cash links
- thehalvening
- 10-print-in-python
- stephenking.com
- Lively PowerfulBird game announcemnt
- Lively PowerfulBird game
- twitter.com/Channel4News
- Random UK post code generator
- Reddit Atlas
- Project Atlas
- Generating Fantasy Maps
- Scrappy fast crawler and spider
- Augur Prediction
- Consensus & Paxos
- kafka & Message Queues
- Awesome blog, full of code
- math-aware search engine
- south korea trending posts on steemit
- github markdown
- crushthestreet.com
- cnblogs.com (chinese)
- 庄周梦蝶
- erlang article at cnblogs
- jvm article at cnblogs
- douban looks awesome
- Government Types
- Jp live musics
- rotten tomatoes
- imdb.com
- geeking.co
- nerdist.com
- soundcloud/subwayssounds
- dystopian
- cyberpunk
- serial experiments of lain
- dystopian novels
- cyverpunky room
- what is dystopia
- lain on kinopoisk
- OMG Ubuntu on Pinterest
- gratuitous-cyberpunk
- awesome-gophers
- clojure in detail (chinese)
we're up all night to get lucky