Icon

NLP - Sentiment Analysis and Topic Modelling

Observing the difference after Preprocessing Steps using Document Viewer node Note on 'Document Vector' node:In order to include the Term Frequenciesin the 'Document Vector' node pleaseuncheck the 'Bitvector' option in theconfiguration window of the node andchoose the term frequency columngenerated in the previous step. Note:We can remove theDocument columnas it won't be usedto build the model. Data Preprocessing Model Building Consider this as thedeployable model. Trainedin a non-productionenvironment, extracted andand promoted to higherinstances Consider this as the trained modeldeployed in production Pre-processing Use this to input thedescriptive text for a new tweetas if via a user interface inproduction Perform all the pre-processing steps on the new data This is for cross validation only. Not really relevantin production as you wouldn't have the classcolumn Retain only the predicted class and view as html Topic Modelling Explore options to combine topics asdocument class itself for furtherenhancements Combine the new documentvector with the one used fortraining/testing retain only the new data so thatthe prediction doesn't happenon all the data Bigger Data Set Read Tweet DataConvert reviewsto documentsFilter outnon-document columnsRemove punctuationRemove wordsrepresentingnumbersRemove small wordsRemove stop wordsTo lower caseStem to the root wordsBoWCompute term frequenciesCreate document vectorExtract labelSplit toTrain/TestComputeperformanceView the DocumentsView the DocumentsRemove the Document columnPruned Decision TreePrediction on Test dataNode 49Node 50Node 51Node 52Node 53Node 54Node 55Node 56Node 57Create document vectorExtract labelRemove the Document columnSplit toTrain/TestComputeperformanceNode 65Node 66Node 67Node 68Node 69Node 71Node 72Node 73Node 74Node 75Node 77Node 80Node 81Node 82Create document vectorExtract labelNode 85Node 86Node 87Node 88Node 89Node 90Node 91Remove the Document columnNode 96Node 97Pruned Decision TreePrediction on Test dataNode 100Node 101Node 102Node 104Node 105Node 106Node 107Node 108Node 109Node 110Node 111Node 112Node 113Node 114Node 115Node 116 CSV Reader Strings to Document Column Filter Punctuation Erasure Number Filter N Chars Filter Stop Word Filter Case Converter Snowball Stemmer Bag Of WordsCreator TF Document Vector Category to Class Partitioning Scorer Document Viewer Document Viewer Column Filter DecisionTree Learner Decision TreePredictor String Cleaner Strings to Document Column Filter Number Filter N Chars Filter Stop Word Filter Snowball Stemmer Bag Of WordsCreator TF Document Vector Category to Class Column Filter Partitioning Scorer IDF StanfordNLPNE Tagger Stanford Lemmatizer Topic Extractor(Parallel LDA) GroupBy t-SNE (L. Jonsson) Tag Cloud(JavaScript) Joiner Scatter Plot Color Manager Table Editor(JavaScript) CSV Reader Bag Of WordsCreator TF Document Vector Category to Class String Cleaner Strings to Document Column Filter Number Filter N Chars Filter Stop Word Filter Snowball Stemmer Column Filter PMML Writer PMML Reader DecisionTree Learner Decision TreePredictor Concatenate Decision TreePredictor Scorer Row Filter(deprecated) Table View(JavaScript) Column Filter CSV Reader Data Explorer Document Viewer CSV Reader Stanford Lemmatizer StanfordNLPNE Tagger StanfordNLPNE Tagger Stanford Lemmatizer Document Viewer Document Viewer Observing the difference after Preprocessing Steps using Document Viewer node Note on 'Document Vector' node:In order to include the Term Frequenciesin the 'Document Vector' node pleaseuncheck the 'Bitvector' option in theconfiguration window of the node andchoose the term frequency columngenerated in the previous step. Note:We can remove theDocument columnas it won't be usedto build the model. Data Preprocessing Model Building Consider this as thedeployable model. Trainedin a non-productionenvironment, extracted andand promoted to higherinstances Consider this as the trained modeldeployed in production Pre-processing Use this to input thedescriptive text for a new tweetas if via a user interface inproduction Perform all the pre-processing steps on the new data This is for cross validation only. Not really relevantin production as you wouldn't have the classcolumn Retain only the predicted class and view as html Topic Modelling Explore options to combine topics asdocument class itself for furtherenhancements Combine the new documentvector with the one used fortraining/testing retain only the new data so thatthe prediction doesn't happenon all the data Bigger Data Set Read Tweet DataConvert reviewsto documentsFilter outnon-document columnsRemove punctuationRemove wordsrepresentingnumbersRemove small wordsRemove stop wordsTo lower caseStem to the root wordsBoWCompute term frequenciesCreate document vectorExtract labelSplit toTrain/TestComputeperformanceView the DocumentsView the DocumentsRemove the Document columnPruned Decision TreePrediction on Test dataNode 49Node 50Node 51Node 52Node 53Node 54Node 55Node 56Node 57Create document vectorExtract labelRemove the Document columnSplit toTrain/TestComputeperformanceNode 65Node 66Node 67Node 68Node 69Node 71Node 72Node 73Node 74Node 75Node 77Node 80Node 81Node 82Create document vectorExtract labelNode 85Node 86Node 87Node 88Node 89Node 90Node 91Remove the Document columnNode 96Node 97Pruned Decision TreePrediction on Test dataNode 100Node 101Node 102Node 104Node 105Node 106Node 107Node 108Node 109Node 110Node 111Node 112Node 113Node 114Node 115Node 116 CSV Reader Strings to Document Column Filter Punctuation Erasure Number Filter N Chars Filter Stop Word Filter Case Converter Snowball Stemmer Bag Of WordsCreator TF Document Vector Category to Class Partitioning Scorer Document Viewer Document Viewer Column Filter DecisionTree Learner Decision TreePredictor String Cleaner Strings to Document Column Filter Number Filter N Chars Filter Stop Word Filter Snowball Stemmer Bag Of WordsCreator TF Document Vector Category to Class Column Filter Partitioning Scorer IDF StanfordNLPNE Tagger Stanford Lemmatizer Topic Extractor(Parallel LDA) GroupBy t-SNE (L. Jonsson) Tag Cloud(JavaScript) Joiner Scatter Plot Color Manager Table Editor(JavaScript) CSV Reader Bag Of WordsCreator TF Document Vector Category to Class String Cleaner Strings to Document Column Filter Number Filter N Chars Filter Stop Word Filter Snowball Stemmer Column Filter PMML Writer PMML Reader DecisionTree Learner Decision TreePredictor Concatenate Decision TreePredictor Scorer Row Filter(deprecated) Table View(JavaScript) Column Filter CSV Reader Data Explorer Document Viewer CSV Reader Stanford Lemmatizer StanfordNLPNE Tagger StanfordNLPNE Tagger Stanford Lemmatizer Document Viewer Document Viewer

Nodes

Extensions

Links

Download

To use this workflow in KNIME, download it from the below URL and open it in KNIME:

Download Workflow
Created by: Ayan Dutta
Created at: 2022-03-28
On NodePit since: 2024-08-20
Last update: 2025-02-18
Created with KNIME version: v5.4.0

Deploy, schedule, execute, and monitor your KNIME workflows locally, in the cloud or on-premises – with our brand new NodePit Runner.

Try NodePit Runner!