Text mining can add considerably to any analysis. In an analysis of only structured data, variables deliberately measures, the focus is narrowed to what is known and expected. Adding text mining of free-form response fields opens the possibilities. It allows analyst to find relationships that weren’t accounted for previously because the analysis is not limited to what the subject matter experts can conceive. Instead, the data tell the story without limitations. Text Mining unlocks a wealth of hidden information and often increases the power and accuracy of predictive models. STATISTICA Text Miner helps to easily uncover valuable information hiding in free form text.
- Turn free-form text into predictor variables: STATISTICA Text Miner indexes terms from free form text and creates numeric indices and components that can be used to broaden your analysis.
- Improve predictive models: Since free form text expands the focus in predictive modeling, new patterns and relationships are uncovered. Incorporating text mining results with data mining and predictive modeling yields better lift.
- Find relationships between terms: Using singular value decomposition results, find key words that occur together frequently in text to reveal how conversations are focused.
- Customize the vocabulary: The tool allows the specification of synonyms, phrases and word to include or exclude to customize the analysis. These lists are specific to the domain and this customization leads to the most interesting results.
- Detect changes over time in what is said in free text: Tracking word frequencies and SVD allows the detection of shifts that indicate changes worth exploring and addressing.