Created
January 15, 2017 11:53
-
-
Save generall/f5fbdab58ecfc78bb6a3eb2e3b6d1d67 to your computer and use it in GitHub Desktop.
Twitter-bot
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Twitter bot for automated detection and re-twitting influential thematic posts. | |
=== | |
Assume that thematic = `programming` | |
##Stage 1 | |
* Prepare examples of programmer's twitter accounts and other ones for representativity | |
* Create graph of twitter accounts which seems to contain programmers thematic | |
* Starting from small amount of truly programmer's accounts collect all of it's followers and following accounts | |
* Add in queue post perspective ones based on | |
* follower count | |
* activity | |
* NLP topic detection (key words) | |
* calculate **PageRank** for programmer's *Graph* | |
###User params in database | |
* `info_extracted` | |
* time of extraction | |
* draft rank | |
* `tweets_extracted` | |
* last tweet id | |
* time of extraction | |
* last tweet time | |
* draft tweet rank | |
* `followers_extracted` | |
* time of extraction | |
* need update? | |
### daemons | |
* info_update deamon | |
* source queue: users with `info_extracted = false` | |
* order: any | |
* API: `client.user` | |
* Action: remove from this queue, add to `tweet_queue`, set `tweets_extracted_count = 0` | |
* tweet load daemon | |
* source queue: users with `info_extracted = true` | |
* order: `[ - tweets_extracted_count, rating]` | |
* API: `c.client.user_timeline` | |
* Action: update `tweets_extracted_count += 1`, update draft tweet rank, update both queues: tweet and followers, add new users to graph from dialog end re-tweets | |
* followers load daemon | |
* source queue: users with `followers_extracted = false, info_extracted = true` | |
* order: `[ rating || draft tweet rating || draft rating]` | |
* API: `client.followers`, `client.followers_ids`, `client.friends`, `client.followers_ids` | |
* Action: Add new users to graph | |
### Graph | |
* All info about users | |
* info about relations | |
* `follow` | |
* `friend` | |
* `retweet` (with count) | |
* `answered` (with count) | |
###Twitter APIs | |
```ruby | |
# get user | |
client.user("@vasnetsov93") | |
# get user`s followers | |
client.followers("@_trean_") | |
# get user`s tweets | |
ts = c.client.user_timeline("@vasnetsov93", count: 100) | |
ts.first.attrs | |
# check limits | |
Twitter::REST::Request.new(tweet_loader.client, :get, 'https://api.twitter.com/1.1/application/rate_limit_status.json', resources: "followers").perform | |
#=> {:rate_limit_context=>{:access_token=>"380767551-Oas6V96LbMdllrbEdVnKoj1ehhskFoYg30wBwSpg"}, :resources=>{:followers=>{:"/followers/ids"=>{:limit=>15, :remaining=>15, :reset=>1483244343}, :"/followers/list"=>{:limit=>15, :remaining=>14, :reset=>1483243574}}}} | |
``` | |
```ruby | |
``` | |
## Stage 2 | |
* Watch posts of perspective peoples for last 24hrs. | |
* If amount of calculated rating is above some threshold, calculated for this account, show this tweet as candidate for reposting. Rating calculation should assume: | |
* Weighted by *Graph* re-tweet and like sum | |
* re-tweet dynamic | |
* Keep collect and watch statistics for selected tweets | |
> Written with [StackEdit](https://stackedit.io/). |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment