Angelina Tzacheva

Abstracts

4 Emotion Classification Using Recurrent Neural Network and Scalable Pattern Mining

Authors: Angelina Tzacheva, Jaishree Ranganathan, Muthupriya Shanmugakani Velsamy, Shamika Kulkarni

Abstract:

Emotions play an important role in everyday life. Analyzing these emotions or feelings from social media platforms like Twitter, Facebook, Blogs, and Forums based on user comments and reviews plays an important role in various factors. Some of them include brand monitoring, marketing strategies, reputation, and competitor analysis. The opinions or sentiments mined from such data helps to understand the current state of the user. It does not directly provide intuitive insights on what actions to be taken to benefit the end user or business. Actionable Pattern Mining method provides suggestions or actionable recommendations on what changes or actions need to be taken in order to benefit the end user. In this paper, we propose automatic classification of emotions in twitter data using Recurrent Neural Network - Gated Recurrent Unit. We achieve training accuracy of 87.58% and validation accuracy of 86.16%. Also, we extract action rules with respect to the user emotion, that help provide actionable suggestion.

Keywords: Twitter, recurrent neural network, gated recurrent unit, emotion mining, actionable pattern mining

Procedia PDF Downloads 1
3 Scalable Action Mining for Recommendations to Reduce Hospital Readmission

Authors: Angelina Tzacheva, Arunkumar Bagavathi, Apurwa Apurwa

Abstract:

Hospital re-admission problem is one of the longtime issues of healthcares in USA. Unplanned re-admissions to hospitals not only increase cost for patients, but also for hospitals and taxpayers. Action mining is one of the data mining approaches to recommend actions to undertake for an organization or individual to achieve required condition or status. Undertaking such actionable recommendations incur some form of cost to users. The actionable recommendation system fails when the recommended actions are cost wise unendurable or non-profitable and uninteresting to the end user. Finding low cost actionable patterns in larger datasets is a time consuming and requires a scalable approach. In this work, we propose a scalable action mining method to recommend hospitals and taxpayers on what actions would potentially reduce patient readmission to hospitals at lowest costs. Most importantly we incorporated graph search methods to extract low cost actionable patterns. We use the Healthcare Cost and Utilization Project(HCUP) databases to evaluate our approach. All our proposed scalable approaches are cloud based and use Apache Spark to handle data processing and to make recommendations.

Keywords: Data Mining, Scalable, action mining, hospital readmission

Procedia PDF Downloads 1
2 Pattern Discovery from Student Feedback: Identifying Factors to Improve Student Emotions in Learning

Authors: Angelina Tzacheva, Jaishree Ranganathan

Abstract:

Interest in (STEM) Science Technology Engineering Mathematics education especially Computer Science education has seen a drastic increase across the country. This fuels effort towards recruiting and admitting a diverse population of students. Thus the changing conditions in terms of the student population, diversity and the expected teaching and learning outcomes give the platform for use of Innovative Teaching models and technologies. It is necessary that these methods adapted should also concentrate on raising quality of such innovations and have positive impact of student learning. Light-Weight Team is an Active Learning Pedagogy, which is considered to be low-stake activity and has very little or no direct impact on student grades. Emotion plays a major role in student’s motivation to learning. In this work we use the student feedback data with emotion classification using surveys at a public research institution in the United States. We use Actionable Pattern Discovery method for this purpose. Actionable patterns are patterns that provide suggestions in the form of rules to help the user achieve better outcomes. The proposed method provides meaningful insight in terms of changes that can be incorporated in the Light-Weight team activities, resources utilized in the course. The results suggest how to enhance student emotions to a more positive state, in particular focuses on the emotions ‘Trust’ and ‘Joy’.

Keywords: Education, Data Mining, emotion, actionable pattern discovery

Procedia PDF Downloads 1
1 Emotion Mining and Attribute Selection for Actionable Recommendations to Improve Customer Satisfaction

Authors: Angelina Tzacheva, Jaishree Ranganathan, Poonam Rajurkar, Zbigniew Ras

Abstract:

In today’s world, business often depends on customer feedback and reviews. Sentiment analysis helps identify and extract information about the sentiment or emotion of the topic or document. Attribute selection is a challenging problem, especially with large datasets in actionable pattern mining algorithms. Action Rule Mining is one of the methods to discover actionable patterns from data. Action Rules are rules that help describe specific actions to be made in the form of conditions that help achieve the desired outcome. The rules help to change from any undesirable or negative state to a more desirable or positive state. In this paper, we present a Lexicon based weighted scheme approach to identify emotions from customer feedback data in the area of manufacturing business. Also, we use Rough sets and explore the attribute selection method for large scale datasets. Then we apply Actionable pattern mining to extract possible emotion change recommendations. This kind of recommendations help business analyst to improve their customer service, which leads to customer satisfaction and increase sales revenue.

Keywords: Data Mining, emotion, Attribute selection, actionable pattern discovery, business data

Procedia PDF Downloads 1