See full project here: https://litersta.com

Complete application code is not loaded to this repository.



Litersta - textual analytics - software leverages statistical algorithms to programmatically locate, and extract, overall document sentiment, word frequencies, and document similarities.

Software Abilities

Similarity Scoring

Similarity scoring is used to locate text documents that are similar to a selected few. A small subset of documents is chosen and each document within the subset is scored based on its similarity to the documents in the population.



Machine Learning

Normalization, categorization, and clustering are used when preparing text for similarity scoring and sentiment analysis.



Word Frequency

The frequency feature is used to locate common words across a selected document set. After the removal of stop words, 5000 words with the highest frequencies are collected from each document. The frequencies from each document are added, and a report that lists the sum of the frequencies is created.



Security

Litersta runs locally, behind your firewall, to make sure that all data is kept safe and secure.



Data Stors and Categories

Data stors are created and are linked to categories. A data stor may be linked to multiple categories; however, a category will never be linked to multiple data stors.



Sentiment Analysis

The text from each document in a data stor category is parsed - positive and negative terms counted - and a report is generated that lists the number of positive and negative terms.



Results