Question regarding both Sentiment Analysis and Forecasting (Times Series) in one fell swoop
wufutura
New Altair Community Member
Hi All!
You really helped me last time I asked a question and really appreciate those of you who jumped in. I have attempted to move on with this bit of knowledge gained and am I trying now to do four things:
1. Use the existing Sentiment Analysis template as a database
2. Incorporate MonkeyLearn into the whole process to presumably aid in processing data when it comes from Twitter on its way to RapidMiner?
3. Perform a times series forecast using the Forecast (Univariate) Operator built into RapidMiner as practice for more adventurous time series forecasting in the future
4. Have applied with the kind folks at Twitter for a real time feed from them, so I can extract real time, rather than static, data once I get #s 1, 2 & 3 above figured out.
(The problems refer visually to the picture included at the top of the page)
So, then I run it and it crunches for just a bit but then returns the following message:
Exception:com.rapidminer.operator.OperatorException
Message: No such configured object of name [edited because this is likely the API token - mschmitz] of MonkeyLearn API Key
Stack trace: (and then a bunch of stuff comes after that that it is literally incomprehensible to me and too long to include here!)
Any thoughts, anyone? I realize I am trying to do two or three things at once herein that may or may not be a good idea, i.e. learning how to do a proper and simple time series forecast and also a sentiment analysis simultaneously....
Please keep in mind I am a rank beginner and need a no-code environment approach to learning. If that isn't already quite apparent!!!!! Yours, Richard
0
Best Answers
-
Good morning, everyone!Once again, I'd like to thank - especially aengler and mschmitz - for their patience and ongoing tutelage. In attempts to employ the above use case I contacted both Twitter and MonkeyLearn. Unfortunately, for reasons that aren't entirely clear to me, the latter (MonkeyLearn) claims that they will ultimately be unable to take me where I want to go.They didn't have any specific alternative provider suggestions other than to advise me to contact Twitter directly, which I did. For my purposes, Twitter feels I should go via AWS.QUESTION: does RapidMiner have the ability to pull from AWS, and, if so, how would I accomplish that? A diagram, if appropriate, with as many specifics you have time for, and friendly/targeted "liberal arts person" prose, as opposed to overtly assumed technical terms, are most helpful in my abilities to understand your intended meanings & explanations.Thanks again/greetings from California, Richard0
-
@wufutura
Hello
Here is a great link from @sgenzer for your question.
https://community.rapidminer.com/discussion/55379/using-aws-image-ml-inside-a-rapidminer-process
I hope this helps
Sara5 -
Hi @wufutura ,if this is a commercial opportunity, then we should go the commercial ways. Our community support is sadly limited, so we cannot deep dive into those issues, since we are doing mainly client/prospect facing work. Please connect with me via email if you want to go down this way. its mschmitz at rapidminer.comBest,MArtin5
Answers
-
Hi @wufutura ,first of all: Cool use case! So the idea is to forecast the opinion on a topic over time? Maybe an Aggregate or moving average operator in between helps to de-noise it.The other thing: Your error stems most likely from the MonkeyLearn operator (the Stack Trace would tell us exactly what happend).Did you create a connection object for MonkeyLearn? It uses the old format, which is described here: https://docs.rapidminer.com/latest/studio/connect/legacy.html .It looks a bit like you just posted your key into the field.I also took the liberty to edit your message a bit. The string/hash you added there looks very much like an API key. I think we do not want to share this publically.Best,Martin
1 -
would go operator by operator per break-points. Most certainly it will fail at a monkeylearn operator. If so ...most likely ... there are several options:Do you have a monkeylearn licenseDo set the options of the operator according to your caseIs the connection to the Monkeylearn API establishedHave you tried to call the Monkeylearn API for yourself via curl to check if everything should work?2
-
Let me try and dissect what you guys mean by some technical terms you are using here. could be a bit of a learning curve for me but one I'm more than willing to climb! thanks, and I may get back to you once I've followed you're advice/guidance/instructions. Thanks so much, Richard
0 -
Good morning, everyone!Once again, I'd like to thank - especially aengler and mschmitz - for their patience and ongoing tutelage. In attempts to employ the above use case I contacted both Twitter and MonkeyLearn. Unfortunately, for reasons that aren't entirely clear to me, the latter (MonkeyLearn) claims that they will ultimately be unable to take me where I want to go.They didn't have any specific alternative provider suggestions other than to advise me to contact Twitter directly, which I did. For my purposes, Twitter feels I should go via AWS.QUESTION: does RapidMiner have the ability to pull from AWS, and, if so, how would I accomplish that? A diagram, if appropriate, with as many specifics you have time for, and friendly/targeted "liberal arts person" prose, as opposed to overtly assumed technical terms, are most helpful in my abilities to understand your intended meanings & explanations.Thanks again/greetings from California, Richard0
-
@wufutura
Hello
Here is a great link from @sgenzer for your question.
https://community.rapidminer.com/discussion/55379/using-aws-image-ml-inside-a-rapidminer-process
I hope this helps
Sara5 -
Thanks, Sara! Question: when I go to the New Connections dropdown area I notice that it gives me an option for Twitter as well as for AWS. Why can't I go to them directly without going through AWS as an intermediary? - Richard
0 -
Hi all - I checked with Twitter, and they say that both Postman and Insomnia are GUI tools and resources that can enable me to connect to the Twitter API and/or access Twitter data.Question: are either or both or these useable within RapidMiner, since a GUI would help right now. I am studying Python but am not yet conversant enough at this point to readily use it to move forward.And, unfortunately, MonkeyLearn says that my ultimate plans cannot be accomplished with their product.Thanks again, Richard0
-
Hi Martin - (mschmitz) - I did not get this question answered and wonder if, since you are the Head of Data Science Services, you might have some insight into it?Here it is again:"Hi all - I checked with Twitter, and they say that both Postman and Insomnia are GUI tools and resources that can enable me to connect to the Twitter API and/or access Twitter data.Question: are either or both or these useable within RapidMiner, since a GUI would help right now. I am studying Python but am not yet conversant enough at this point to readily use it to move forward.And, unfortunately, MonkeyLearn says that my ultimate plans cannot be accomplished with their product.Thanks again, Richard"
0 -
Martin - any further thoughts on this question re a GUI i could use? Richard
0 -
Martin - Do I now need anything else other than some
- severe hand holding to get AWS and Twitter on to RapidMinder now? - as they have,
- both just put me on their systems...
If my POC works I will have quite a bit of business for you, and for them - of course. Stay well, stay healthy, Richard0 - severe hand holding to get AWS and Twitter on to RapidMinder now? - as they have,
-
Hi @wufutura ,if this is a commercial opportunity, then we should go the commercial ways. Our community support is sadly limited, so we cannot deep dive into those issues, since we are doing mainly client/prospect facing work. Please connect with me via email if you want to go down this way. its mschmitz at rapidminer.comBest,MArtin5
-
it is, yes. how much are we talking to get this going?
0 -
sorry - didn't see the request to communicate via email. my bad. richard wufutura@protonmail.com415-816-2737(i don't are if anybody else sees this, actually)0
-
interested in pursuing the correct mode of communication er this topic.
0