sentiment | Sentiment analysis using machine learning techniques | Machine Learning library

Good day fam. I have this sentiment analysis API from Paysify, which returns a JSON output but all of a sudden it is now returning NULL as output. Please I help here, I have been trying to fix this for the past 4 hours, no headway.

I want to force the Huggingface transformer (BERT) to make use of CUDA. nvidia-smi showed that all my CPU cores were maxed out during the code execution, but my GPU was at 0% utilization. Unfortunately, I'm new to the Hugginface library as well as PyTorch and don't know where to place the CUDA attributes device = cuda:0 or .to(cuda:0).

The code below is basically a customized part from german sentiment BERT working example

I am trying to perform a sentiment analysis based on product reviews collected from various websites. I've been able to follow along with the below article until it gets to the model coefficient visualization step.

https://towardsdatascience.com/how-a-simple-algorithm-classifies-texts-with-moderate-accuracy-79f0cd9eb47

When I run my program, I get the following error:

I'm running several machine learning models to find the one which the highest accuracy score, however, all the accuracy scores are the exact same. I performed NLP on social media text and I'm training my models to tag sentiment based on sentiment determined from NLTK.

I'm using the same training and test sets, but I've done this method before in the past and received different scores on different models. Why are all of mine the same? Am I overfitting perhaps?

Here is my code where I'm splitting and training:

I'm just starting to learn python and I'm having problem with displaying two widget at the same time. I was trying to create a Text area where I will put a string then submit this string. This string will be use in SentimentIntensityAnalyzer to check the sentiment.

My goal now is to display the two widget using ipywidgets below is my code

I am currently trying to setup a gcloud appspot domain. I am consistently getting this error when I attempt to open up my appspot project on the browser:

Error: Not Found The requested URL / was not found on this server.

Here is my code. How can I fix this?

app.yaml -

I have been executing a sentiment analysis, and have returned the positive and negative outcomes to my pd.DataFrame in the following manner.

However, when I want to split the sentiment column (which consists of Polarity and Subjectivity), I get the following error:

ValueError: Columns must be same length as key

I have tried multiple approaches:

• str.split
• str.extract
• rounding

With the rounding approach I get an error mentioning the float NaN's could not be multiplied. So I suppose that there is a NaN in there somewhere. However, when I look for NaN's I get this answer:

I am learning about text mining and rTweet and I am currently brainstorming on the easiest way to clean text obtained from tweets. I have been using the method recommended on this link to remove URLs, remove anything other than English letters or space, remove stopwords, remove extra whitespace, remove numbers, remove punctuations.

This method uses both gsub and tm_map() and I was wondering if it was possible to stream line the cleaning process using stringr to simply add them to a cleaning pipe line. I saw an answer in the site that recommended the following function but for some reason I am unable to run it.

I'm using Plotly to plot some sentiment data yet I have the issue that one character has much more lines than the other so it seems that we cannot compare the two. I'm unsure how to describe for this to find the solution so I hoped that perhaps someone will understand what I'm referring to. Here are images:

We cannot really compare orange's negativity with blue's negativity, or can we? Should we have the orange have 600 as the label, it being much smaller than blue or is this correctly scaled?

I'd like this so we can say that for example, blue is less positive than orange.

I'm sorry for any confusion. Thank you.