Skweak for text classification about skweak HOT 5 CLOSED

Well, the sentiment analysis code is an example of text classification. More generally, you can do classification simply by implementing labelling functions that create spans covering the entire text (that is, from 0 to len(doc)), and then running the aggregation. Then only thing you have to remember is to set the argument sequence_labelling to False.

from skweak.

Thank you.

from skweak.

Hi, @plison I have a quick question,
To find the sentiment of sentences we many many libraries from markets like TextBlob, NLTK, Transformers, Flair...etc.
so when we don't have labeled data to train the sentiment analysis model we can get the labels from those libraries. and then we can train our model.
and how can I get labels to my domain data?
for ex: I have text like "my printer is not working" and I want to label it as "Hardware problem".
How can I achieve this with Skweak? Is there any demo code for that kind of labeling?

from skweak.

Well, to label your data with domain-specific labels (such as HardwareProblem), you can either write some heuristic rules (such as a rule stating that, if the word "printer" occurs in the sentence, it is likely to be a hardware problem), or have crowdsourcing annotators label those by hand. If you have several of those heuristics / annotators applied to the same texts, you can then use skweak to aggregate their predictions and get a single annotation layer that merges all of them together. Then, the only remaining step is to learn a classical neural classification model (like a language model with a text classification head) based on those aggregated annotations.

from skweak.

from skweak.