follow source

[lavalamp]

hey there just want to ask if you want to follow the "users who interact to a tweet", what option will i choose? is it "tweetslikedbyuser"? or "tweet Interactions" (what the difference of those 2?)

I mean, I want to scrape the user who likes the tweet. what the option to choose?


thanks in advance

[martin@rootjazz]

"tweetslikedbyuser"? or "tweet Interactions" (what the difference of those 2?)

tweetslikedby user, are the tweets that user has liked

Tweet inteactions this is what you want.

It's the likers / replies / retweeters of a specific tweet

I mean, I want to scrape the user who likes the tweet. what the option to choose?

if you just want the likers, then use the step LIKERS



Regards,
Martin

[lavalamp]

thanks

[lavalamp]

so what I did is

Input the "user URL" in the input
choose "user tweet" then choose "likers" in options box

what I intended the bot will do is "search for the user tweet and scrape who likes the tweet"

am i doing it right?

And also, will this settings hit the twitter RATE LIMIT? as MAX SCRAPE RESULT in settings tab is 5000. do i need to reduce it?




(Didn't knew you can choose 2 in options as I only select "followersOf" before)

[martin@rootjazz]

Input the "user URL" in the input
choose "user tweet" then choose "likers" in options box

as user tweets takes an input of USER who don't need the ID / URL / USER step, doesn't make any difference, but it's not required

am i doing it right?

you are always just try it, for test, 5 results, see what happens

And also, will this settings hit the twitter RATE LIMIT? as MAX SCRAPE RESULT in settings tab is 5000. do i need to reduce it?

maybe, the new twitter rate limits are new, I don't really know about them. Better to specify the results you want in the action and via the custom search step results wanted, than relying on the protection on the Settings tab. Also, if you don't need 1million results, don't try and scrape them just "because you can". If you can only feasibly use 500 results, only scrape 500, not 5000 etc


Also, run tests. If you have an idea for a search, run a sample of 100, what are the results like, check them out, are they quality results. I had a case of a customer tell me the program is crap, they have follow 10,000 people and got zero new followers. Checking out who they had followed (a Followers_Of search of an "influencer" it turns out that influencer had purchased 10,000 fake followers and this customer had just followed 10k fake accounts. A quick look at who they would have been following would have shown this. Following a sample, 100, then seeing if you get any follow backs would have shown they got 0% follow backs.

So work smarter, not harder, it's about results, not numbers. If you can get better results processing 100 items, that's better than worse results processing 10k items

[lavalamp]

I am testing it now with this "user tweet - likers" and getting the result I wanted (though I'm afraid I might hit the rate limit).



I have a USER URL as my source and if I use "likers or "tweet interactions" it will return with this "FAILED: 308576766378577 : Couldn't extract tweet ID:". I think its looking for "tweet URL" as a source.

Is there a way to use "tweet interaction" and will use USER URL as a source/input

What I want the bot will do is to choose just 1 random tweet (with the "USER URL as my source) then scrape the user who likes the tweet. Is there an options for that?


Additional question
If there's a remaining items after the follow action, will the bot discard it or save it for the next run?

[martin@rootjazz]

I am testing it now with this "user tweet - likers" and getting the result I wanted (though I'm afraid I might hit the rate limit).

I've known a couple of users to hit the limits, nothing wrong with that, especially as they are quite restrictive now. Twitter just returns a 429 / rate limit response and doesn't return data for the searches.

I have a USER URL as my source and if I use "likers or "tweet interactions" it will return with this "FAILED: 308576766378577 : Couldn't extract tweet ID:". I think its looking for "tweet URL" as a source.

Correct, think about. You cannot "like" a profile. A profile has "no likers".

Watch the custom search tutorial, it's all explained:
Twitterdub custom search tutorial
- Explains how to build a search and specify how many results per step (optional)

https://www.youtube.com/watch?v=D1W2KnNciPI

Single click a search step option and it tells you the INPUT and the OUPUT.

Click "likers" or "interactions" and it tells you, the input required is a Tweet. You are passing a profile to that step so it doesn't work

Is there a way to use "tweet interaction" and will use USER URL as a source/input

Yes, you need to get the profiles tweets, then get the likers / interactions from the tweet.

so:
input: @target

Steps:
USER TWEETS (10)
LIKERS (10)

Which means, get 10 of the most recent tweets, then get 10 likers from each, giving a max result set of 100 (10x10). You can change these values to whatever you like. All explained in the tutorial

What I want the bot will do is to choose just 1 random tweet (with the "USER URL as my source) then scrape the user who likes the tweet. Is there an options for that?

Why limit to one tweet? Just use USER TWEETS step with a limit to avoid over scraping (i.e. 10) then get the Likers. If you set your Max Results to a value, when you have that many results, the searching stops.

Additional question
If there's a remaining items after the follow action, will the bot discard it or save it for the next run?

If you tell the program to follow 100 items, it will only scrape 100 items.

But if the action does end early with unprocessed items, they will be saved, the logs should say where the items are saved to

[lavalamp]

thanks for the reply
I watched the tutorial and have a clearer idea now. (didn't knew you have a new tutorials, cant hear the audio on your old tutorial)

a few question. if I use popular tweets/ user tweet/ recent tweet. does the bot detects and ignore the ads/promoted tweet?

and does the bot choose randomly on popular tweets/ user tweet/ recent tweet

[martin@rootjazz]

thanks for the reply
I watched the tutorial and have a clearer idea now. (didn't knew you have a new tutorials, cant hear the audio on your old tutorial)

Once I have found the time to redo all tutorials, I'll delete the old tutorials, just finding the time to make tutorials is difficult.

a few question. if I use popular tweets/ user tweet/ recent tweet. does the bot detects and ignore the ads/promoted tweet?

Yes ads removed (or should be) With the searches all being updated, some ads might have been missed but I think I got them all

and does the bot choose randomly on popular tweets/ user tweet/ recent tweet

No, it's results order.

So for recent / user tweets they are date order, so if you want a 100 you get the 100 most recent items. There should be an action tab order to then randomise the results.