Classifying 200k articles in 7 hours using NLP
1–10 of 32 posts
Re: Classifying 200k articles in 7 hours using NLP
#2The title makes it sound like they talk about how they did it so efficiently.
But all the info we get about that is
"The last step was to combine the four binary models into one multiclass model, as explained in the previous section, and use it to classify 1M new documents automatically. To do this, we simply went on the UI and uploaded a new list of documents."
Great intro to NLP article, but very light on the actual implementation details and dataset.
Re: Classifying 200k articles in 7 hours using NLP
#3The title makes it sound like they talk about how they did it so efficiently. But all the info we get about that is "The last step was to combine the four binary models into one multiclass model, as explained in the previous section, and use it to classify 1M new documents automatically. To do this, we simply went on the UI and uploaded a new list of documents." Great intro to NLP article, but very light on the actua…
Given that is of interest, I'll do a follow up on the implementation details and dataset!
Re: Classifying 200k articles in 7 hours using NLP
#4The title makes it sound like they talk about how they did it so efficiently. But all the info we get about that is "The last step was to combine the four binary models into one multiclass model, as explained in the previous section, and use it to classify 1M new documents automatically. To do this, we simply went on the UI and uploaded a new list of documents." Great intro to NLP article, but very light on the actua…
Given that is of interest, I'll do a follow up on the implementation details and dataset!
very interested!
Re: Classifying 200k articles in 7 hours using NLP
#5The title makes it sound like they talk about how they did it so efficiently. But all the info we get about that is "The last step was to combine the four binary models into one multiclass model, as explained in the previous section, and use it to classify 1M new documents automatically. To do this, we simply went on the UI and uploaded a new list of documents." Great intro to NLP article, but very light on the actua…
Given that is of interest, I'll do a follow up on the implementation details and dataset!
Same, very interested. Cool intro to NLP, but those are easy to come by. Some implementation details and perhaps some code to reproduce would be amazing.
Re: Classifying 200k articles in 7 hours using NLP
#6I wonder what classifier was used (assume neural network-based, given the figures), and how that compares to a simple baseline that uses bag-of-words, such as a linear model or naive Bayes. The examples look easy enough to be classified by matching keywords.
Re: Classifying 200k articles in 7 hours using NLP
#7I wonder what classifier was used (assume neural network-based, given the figures), and how that compares to a simple baseline that uses bag-of-words, such as a linear model or naive Bayes. The examples look easy enough to be classified by matching keywords.
We used a shallow neural network. The main challenge at Sculpt wasn't the modelling part, but rather the whole UX (active learning to speed up the training process, explainability, easily get predictions, etc). So it's true that a relatively simple model performs well, and actually having an efficient / shallow network also helps make the active learning pipeline fast for the user.
Re: Classifying 200k articles in 7 hours using NLP
#8I wonder what classifier was used (assume neural network-based, given the figures), and how that compares to a simple baseline that uses bag-of-words, such as a linear model or naive Bayes. The examples look easy enough to be classified by matching keywords.
We used a shallow neural network. The main challenge at Sculpt wasn't the modelling part, but rather the whole UX (active learning to speed up the training process, explainability, easily get predictions, etc). So it's true that a relatively simple model performs well, and actually having an efficient / shallow network also helps make the active learning pipeline fast for the user.
Thanks! The dashboard looks like a nice piece of work indeed.
Re: Classifying 200k articles in 7 hours using NLP
#9The title makes it sound like they talk about how they did it so efficiently. But all the info we get about that is "The last step was to combine the four binary models into one multiclass model, as explained in the previous section, and use it to classify 1M new documents automatically. To do this, we simply went on the UI and uploaded a new list of documents." Great intro to NLP article, but very light on the actua…
Given that is of interest, I'll do a follow up on the implementation details and dataset!
Would also love if you could share code & data!
Re: Classifying 200k articles in 7 hours using NLP
#10Is Sculpt AI available for public use? I see sculptintel.com is down.