Using Dcipher Studio
A self-service text analytics toolkit, equipped with all the necessary data sources, algorithms, and AI models to design and automate the custom workflow that you need
By Tomas and 1 other2 authors70 articles
Getting started
Get started with Dcipher Studio in minutes.
- Setting up a free trial accountTo get started with Dcipher Analytics, set up a free trial account.
- Creating a projectCreate your first project to get started with Dcipher Analytics.
- Importing data from the fileImport data from single or multiple files in any of a number of formats, including JSON, TXT, CSV, TSV, Excel, and PDF.
- Importing data from social mediaUse connectors to blogs, forums, Twitter, Facebook, Instagram, and YouTube to import data from these channels directly into your project.
- Importing data from news mediaAccess news articles from 100+ countries and import them directly into your project
- Importing data from MiroImport Miro boards into Dcipher Analytics and automatically organize Miro notes based on their content
- Importing data from SurveyMonkeySimplify your survey analysis by using SurveyMonkey integration on Dcipher Analytics
- Getting familiarized with the dataGet a visual overview of the data before starting to work with it.
Tips and tricks
Discover ways to get most out of Dcipher Studio and learn about the platform's basic functionality.
- Using workbenchesWorkbenches provide windows into the data to help interpret and understand it, as well as options for working the data.
- Applying operationsOperations are used for processing data in Dcipher. Here is how to use them.
- Understanding operations pipelinesDcipher tracks the workflow by constructing pipelines of operations based on user interactions with the data.
- Triggering operations via drop zonesWorkbench drop zones provide the most relevant processing options for your data and trigger operation sequences based on your objective.
- Selecting data in different workbenchesLearn the shortcuts for selecting data in different workbenches to efficiently apply operations and filters on subsets of data.
Data prep
Get your data ready through data prep: dataset joins, sampling, field conversions, and more.
Preprocessing & cleaning
Everything you need to clean and preprocess your text data before getting started with analysis or model training.
- Near duplicates removalDeletes almost duplicates texts based on the similarity threshold set.
- The Preprocessing Wizard: Fast text cleaning & preprocessingClean and preprocess text through outlier removal, deduplication, language filtering, text cleaning, text splitting, and content extraction.
- Semantic filtering & disambiguationFilter text data based on meaning rather than keywords to account for the ambiguity and richness of natural languages.
- Case conversionThe Change Case operation converts the case of text, with options for lowercasing, uppercasing, and various mixed-case formats.
- Duplicate removalThe Delete Duplicates operation deletes duplicate values.
- Pattern matching & replacementThe Replace Pattern operation replaces a user-specified pattern in text with a user-specified text string.
- Text cleaningThe Clean Text operation provides options for cleaning elements from text, such as URLs, @ tags, hashtags, punctuation, tabs, and emojis.
- Text length measurementThe Calculate Text Length operation counts the number of letters, digits, spaces, other characters, words, and sentences in text.
Text segmentation & tokenization
Split your text into manageable pieces through text segmentation and tokenization.
- Splitting text into sentencesThis operation splits text into sentences while paying attention to language-specific syntax rules.
- Tokenization: splitting text into words and phrasesThe Tokenize & Tag operation splits text into meaningful units in the form of words, phrases, and named entities, tagged by part-of-speech.
- Smart segmentationThe Segment Text operation splits text into segments based on semantic similarity as well as references between adjacent sentences.
- Pattern-based text splittingThe Split by Pattern operation splits text by pattern, for example into sentences or paragraphs.
Content extraction
Extract relevant information from text: content, keywords, quotations, dates, quotes, and more.
- Keyword-based content extractionExtract Content by Keyword extracts sentences containing specified keywords as well as a user-specified number of surrounding sentences.
- Keyword extractionThe Extract Keywords operation identifies and extracts a set of keywords that best describe each input text.
- Quotation extractionUse the Extract Quotations operation to extract quotations from text.
- Date extractionExtract Dates is an operation that identifies and extracts dates from text.
- Property extraction by patternExtract Properties by Pattern is used to extract properties from text based on user-specified patterns in the form of regular expressions.
- Duplicate extractionThe Extract Duplicates operation extracts duplicate values.
Text enrichment
Enrich your text data through state-of-the-art NLP and machine learning models: sentiment, topics, concepts, translation, and more.
- Coreference resolutionFinds all expressions that refer to the same entity in a text.
- Entity extraction & linkingDetects entities and links them to entities stored in knowledge bases.
- Score textAssigns numerical values to texts based on syntactic or semantic similarity
- Training and deployment of custom text classifiers with Dcipher AnalyticsTrain custom text classifiers on unlabeled or partially labeled data using the Active Learning approach.
- Classify text (zero-shot)Predicts labels using BART-based zero-shot classifiers without requiring training a text classifier beforehand.
- Tag by ruleTags values in the selected field based on the defined keyword-based boolean criteria.
- Sentiment analysisThe Analyze Sentiment operation analyzes the sentiment of input text through deep learning or sentiment lexicons.
- EmojizationThe Emojize operation interprets the text and outputs the emojis that best capture the emotional nuances expressed.
- Topic modelingUse Detect Topics to identify topics in large volumes of unlabeled text.
- Concept detectionThe Detect Concepts operation finds the underlying concepts in the input text.
- Language detectionThe Detect Language operation detects the language of each input text and outputs the corresponding ISO 639-1 (2-letter) language code.
- Text-level machine translationThe Translate Text operation uses Google Translate to machine translate the input text.
- Word-level machine translationTranslate Words translates between any pair of 58 available languages, enabling multi-language analysis without full text-level translation.
- Knowledge extractionEnrich your data by extracting knowledge in the form of triplets
- Third-party NLP servicesProvides access to third-party text enrichment services such as categorization and emotion recognition.
Vector transformations
The tools you need to mathematically represent and work with text as vectors.
- Text vectorizationThe Vectorize Text operation transforms text into mathematical vector representations.
- Vector reductionReduces the dimensionality of a multidimensional vector space, for example used to represent text or words, onto a lower dimensional space.
- Vector classification"Classify vectors" operation enables training Machine Learning classifiers and getting predictions from the trained models
- Vector regression"Apply regression on vectors" operation enables training Machine Learning regressors and getting predictions from the trained models
Numerical operations
The numerical operations you need for your text analysis.
- Pivoting data for descriptive statistics in the Table ViewGroup and aggregate data to generate descriptive statistics through drop zones in the Table View.
- Applying Pythonic formulaThe Apply Pythonic Formula operation applies a custom formula using Python syntax.
- Discretization"Discretize" operation transforms continuous values into discrete counterparts
- Normalizing numeric values"Normalize" operation adjust values measured on different scales to a notionally common scale.
Visual exploratory text analysis
Spot and explore patterns in your text data bottom-up, top-down, and through iterative approaches.
- Creating a word cloud from textGenerate a word cloud to get a quick overview of the content of your text.
- Mapping word clusters through contextual word cloudsCluster words based on co-occurrence to generate a contextual word cloud, providing a visual map of topics.
- Exploring topics through foam chartsUse foam charts to visually explore topics in text.
- Document landscapingOrganize texts based on their semantic similarity to visually explore and identify themes in text.
- Finding overrepresented valuesFind out what is over-represented in a subset of the data compared to the whole.
Charts for data visualization
Visualize your data as charts.
- Creating a bump chartUse bump charts to visualize change over time.
- Creating a matrixCalculate relationships between two sets of values and display them in grid format using the Matrix Chart.
- Creating a bar chartUse bar charts to display categorical data and corresponding quantities.
- Creating a scatter plotUse a scatter plot to display data along two dimensions.
Export options
Export analysis results from Dcipher Studio to other applications
Automation
Automate your operations and save time on text analytics
Use cases
Tutorials for specific text analytics use cases, from social media mining to analysis of free-form text responses.