0 ×

01_​Guided_​Labeling_​for_​Document_​Classification

Workflow

Guided Labeling for Document Classification
This workflow defines a fully automated web based application that will label your data using active learning and weak supervision. The workflow was designed for business analysts to easily go through documents to be labeled in any number of classes. In each iteration the user labels more documents and the model is trained using the already labeled instances. With every new iteration, the model proposes documents based on a exploration vs exploitation approach. Once the user is happy with the overall potential falling below a certain value, they can exit the loop and export the model to label the remaining instances. Additionally the workflow lets the user defines rules to label instantly a portion of the dataset with a certain condition. These rules provide weak signals for the weak supervision model training. Rules can be updated at any iteration. This workflow is made to be deployed on KNIME WebPortal via KNIME Server.
active learninghuman-in-the-loopguided analyticsdocument classificationsentiment analysistopic detectiontag cloudtopicactivelearninguncertaintyentropysamplinguncertainty samplingmovie reviewshighlitingexplorationexploitationlabel densitypotentialweak supervisionsnorkelgenerative modeldiscriminative modelweak signalsweakruleslabeling functionhuman-in-the-loop
Guided Labeling for Document ClassificationThis workflow defines a fully automated web based application that will label your data using active learning and weak supervision.This workflow is made to be deployed on KNIME WebPortal via KNIME Server. To test the Guided Analytics application on KNIME Analytics Platform:- Right click the Label component and "Execute and Open Views"- Follow the in-view instructions- After saving your interactions, right click Active Learning Loop End and "Step Loop Execution"- Open the Label Component view again to see the second iteration of the Human-in-the-Loop The Process Step by Step1. Upload your documents and enter / upload the labels you want to use.2. Define the rules to label your data with Weak Supervision.2. Start labeling your data in Active Learning loop and update the rules if necessary.3. Monitor overall potential as you provide more labels4. When the overall potentials falls below a desired amount, exit the loop5. Download the model and the labels, and visualize the results Show user currentpredictions and ask formore labels. Allow user to download themodel trained on all thelabeld data. ExplorationPotentialreduce density top : new rulesbottom : new labelsExploitationport 0 : labeledport 1: not labeledport 0 : termsport 1: docstop: new iteration labelsbottom : labeled + unlabeledtop 50 docuentsby PotentialDensity Scorer Exploration/ExploitationScore Combiner Active LearningLoop Start Active LearningLoop End Density Updater Label Entropy UncertaintyScorer Deploy Initialize / Train Classifierwith Available Labels Upload Text Preprocessing Pre-processing Post-processing Top k Selector Graph DensityInitializer Rules Update Rules Guided Labeling for Document ClassificationThis workflow defines a fully automated web based application that will label your data using active learning and weak supervision.This workflow is made to be deployed on KNIME WebPortal via KNIME Server. To test the Guided Analytics application on KNIME Analytics Platform:- Right click the Label component and "Execute and Open Views"- Follow the in-view instructions- After saving your interactions, right click Active Learning Loop End and "Step Loop Execution"- Open the Label Component view again to see the second iteration of the Human-in-the-Loop The Process Step by Step1. Upload your documents and enter / upload the labels you want to use.2. Define the rules to label your data with Weak Supervision.2. Start labeling your data in Active Learning loop and update the rules if necessary.3. Monitor overall potential as you provide more labels4. When the overall potentials falls below a desired amount, exit the loop5. Download the model and the labels, and visualize the results Show user currentpredictions and ask formore labels. Allow user to download themodel trained on all thelabeld data. ExplorationPotentialreduce densitytop : new rulesbottom : new labelsExploitationport 0 : labeledport 1: not labeledport 0 : termsport 1: docstop: new iteration labelsbottom : labeled + unlabeledtop 50 docuentsby PotentialDensity Scorer Exploration/ExploitationScore Combiner Active LearningLoop Start Active LearningLoop End Density Updater Label Entropy UncertaintyScorer Deploy Initialize / Train Classifierwith Available Labels Upload Text Preprocessing Pre-processing Post-processing Top k Selector Graph DensityInitializer Rules Update Rules

Download

Get this workflow from the following link: Download

Resources

Nodes

01_​Guided_​Labeling_​for_​Document_​Classification consists of the following 811 nodes(s):

Plugins

01_​Guided_​Labeling_​for_​Document_​Classification contains nodes provided by the following 15 plugin(s):