Purpose of the Document

This is a demonstration of what is possible to extract through Emotion Analysis of Text available in a File.

This demonstration is specifically focused on the HR Department of the Tools and Solutions to highlight how the candidate selection process could be enhanced by using Text Mining.

The program takes less than 30 seconds to generate the report for a Resume containing 10 pages of text.

About the Developed Program

This is a prototype which could be developed into a full blown system in 2 man week. The developed system would take a text file as an input and produce the analysis as an output. The program would be available through the Internet. However, the deployment site would be available to General Public as well.

How to use the system?

The program requires a text file as an input.

So, the Resume of the Candidats need to be obtained in a format from which a Text File can be generated. For example, if the Candidate’s Resume is available in Word Document, the document could be saved as Text File to obtain the Text format of the Resume.

Through the GUI available through any browser, the Text file can be uploaded.

Once the Text File is uploaded, the program will generate the analysis. The analysis can be viewed on the browser.

The analysis can be downloaded as a PDF file.

Please see the following sections for the different output produced by the system. The system will not draw conclusions from the findings for each individual cases. The inferences will have to be drawn by the reader of the report.

MOST COMMON WORDS IN THE TEXT

This section displays the list of most common words used in the text. The nature of words used indicates the areas of maximum exposure and areas of expertise.

WORD CONTRIBUTION TO SENTIMENTS

This section highlights the words which have contributed to the various emotions extracted in the text. The emotions extracted is as per the lexicon used. The lexicon assigns a emotion to the words as per generic computation preset in the knowledge base. While evaluating this graph, specific context needs to be kept in mind before drawing conclusions.

## Selecting by sentiment

POSITIVE & NEGATIVE SENTIMENTS

This graph indicates the volume of positive and negative sentiments used through different words used in the text. Higher volume of GREEN colour in the graph indicates positivity in the text, while a higher volume of RED indicates negativity in the text.

SENTIMENTS

This graph indicates the prevalent emotions in the text.

SENTIMENT SCORE

This graph display the different words that have contributed to the different emotions evaluated in the text. The same word can contribute to different emotions based on the context it is used in.

WORD CLOUD

The Word Cloud displays the 250 most used words in the text.

The size of the words indicates the frequency of the use of the individual words. The larger the size of the word, the more frequently the word has been used.

The colours indicates the different emotions in which context the words have been evaluated by the program.

This graphic displays at a glance the prevelant experience and exposure of the candidates.

## Joining, by = "word"

BIGRAM ANALYSIS

This section provides an analysis of the BIGRAMS used in the text. BIGRAMS are combination of 2 continuous words used in the text. BIGRAMS indicates the common associations made in experience and exposure.

This graph displays the most frequently used BIGRAMS.

This graphic provides a pictorial view of the most frequently used BIGRAMS.

TRIGRAM ANALYSIS

This section provides an analysis of the TRIGRAMS used in the text. TRIGRAMS are combination of 3 continuous words used in the text. Just like BIGRAMS, TRIGRAMS indicates the common associations made in experience and exposure.

This graph displays the most frequently used TRIGRAMS.

—— THE END ——