O slideshow foi denunciado.
Seu SlideShare está sendo baixado. ×

Ontology based sentiment analysis

Ontology based sentiment analysis

Baixar para ler offline

I created this presentation to present my research work to the committee. My research was on extracting tweets and analyzing it with an previously created ontology model. The results of the ontology model will help in identifying the domain area of the problem for which use had shared negative sentiments on tweeter. This system along with the ontology model developed for Postal service domain. The next step in research will be to generate automated responses on twitter to the user who shares negative sentiments.

I created this presentation to present my research work to the committee. My research was on extracting tweets and analyzing it with an previously created ontology model. The results of the ontology model will help in identifying the domain area of the problem for which use had shared negative sentiments on tweeter. This system along with the ontology model developed for Postal service domain. The next step in research will be to generate automated responses on twitter to the user who shares negative sentiments.

Mais Conteúdo rRelacionado

Ontology based sentiment analysis

  1. 1. ONTOLOGY-BASED SENTIMENT ANALYSIS MODEL By Pratik Thakor Department of Computer and Information Science Advisor: Dr. Sreela Sasi May 2, 2014
  2. 2. OVERVIEW  Introduction  Problem  Background Research  Proposed Solution  Architecture  Results  Conclusion
  3. 3. INTRODUCTION  Social media connects organizations and customers. i.e. Twitter, Facebook and Google+.  Use of social media:  Organization  Get product feedbacks  Promote brand value  Directly connect with customers.  Customer  Get product updates  Build and connect with product user community  Share experience
  4. 4. PROBLEM  Organizations:  Read direct feedbacks  Generate report of satisfaction/dissatisfactions  Communicate  NO Interactive communication for user’s complaint on social media  A system is needed  Can extract social media content & analyze  Identify the reason for problem  Generate the response on the social media platform
  5. 5. BACKGROUND RESEARCH  Sayed Zeesan Haider, “Ontology-based sentiment analysis case study”, a case study for Master degree project, University of Skovde, pages 05-67, 2012.  Built cell phone feature-based ontology model  Analyzed the customer review  K.M Sam and C.R. Chatwin, “Ontology-Based Sentiment Analysis Model of Customer reviews for Electronic Products”, Proceedings of International Journal of e- Business, e-Management and e-Learning.  Built the customer satisfaction model
  6. 6. BACKGROUND RESEARCH  Tim Finin, Li Ding and Lina Zou “Social Networking on the Semantic Web”, Learning Organization Journal, special issue on Ubiquitous Business Intelligence, Miltiadis Lytras et al, 2005.  Ontology-based intelligent application  Natalya F. Noy and Deborah L. McGuinness “A guide to creating your first ontology”, Stanford University.  Ontology building
  7. 7. PROPOSED SOLUTION  An ontology-based sentiment analysis model and an automated response generator system.  Architecture of the model - three processes  Ontology model creation process  Sentiment analysis with ontology model (Identifying the associated problem with the content)  Automated response generator
  8. 8. ARCHITECTURE PROCESS 1
  9. 9. ARCHITECTURE PROCESS 2
  10. 10. ARCHITECTURE PROCESS 3
  11. 11. ARCHITECTURE - MODULES  Data extraction: Extract data from Twitter  GATE software: Extract information like nouns and verbs from the content  Protégé software: Build ontology model and to query the model  Ontology model: Consists class, subclass, objects, object properties  SentiStrength2: Identify positive and negative sentiments tweets.  SPARQL query language: Query the ontology model and retrieve the information
  12. 12. SAMPLE TWEETS
  13. 13. GATE SOFTWARE OUTPUT
  14. 14. GATE SOFTWARE FILE OUTPUT
  15. 15. SAMPLE ONTOLOGY MODEL
  16. 16. QUERY BUILDING
  17. 17. INFORMATION RETRIEVAL FROM ONTOLOGY MODEL
  18. 18. SENTISTRENGTH2 RESULTS
  19. 19. CONCLUSION  “We can develop a system to analyze negative content being shared on social media platform and try to find out problem associated with it. After understanding the problem, it is possible to generate predefine reply for it on social media.”  This model will help in building foundation for further research on the use of ontology for sentimental analysis.
  20. 20. REFERENCE  Sayed Zeesan Haider, Ontology-based sentiment analysis case study, University of Skovde, pages 05-67, 2012  K.M. Sam and C.R. Chatwin, Ontology-based Sentiment Analysis Model of Customer reviews for Electronic Products, Proceedings of International Journal of e-Business, e-Management and e- Learning, Vol. 3, No. 6, December 2013  Larissa A. de Freitas and Renata Vieira, Ontology- based Feature Level Opinion Mining for Portuguese Reviews, PUCRS FACIN, Porto Alegre, Brazil, 2013
  21. 21. REFERENCE  Bing Liu, “Sentiment Analysis and Subjectivity”, from Handbook of Natural Language Processing, Second Edition, (editors: N. Indurkhya and F. J. Damerau), 2010  Matteo Baldoni, Cristina Baroglio, Viviana Patti and Paolo Rena, “From Tags to Emotions: Ontology-driven Sentiment Analysis in the Social Semantic Web”, Universit`a degli Studi di Torino, 2010  Natalya F. Noy and Deborah L. McGuinness “A guide to creating your first ontology”, Stanford University

Notas do Editor

  • Introduction --- (No background, merge them) social media connect organization and customers i.e. Twitter, Facebook etc
    --- users comment about product (From customer’s perspective)
    --- customer’s response monitor
    --- various surveys, get product feedbacks, to directly connect with customer, maintain brand value(to promote their brand)
    Problem --- Read user’s feedbacks and rare case communicates via social media platform, generate report of satisfaction/dissatisfactions.
    --- fail to attend user’s complaints on this platforms
    --- organization can use social media to provide customer
    Previous research --- (Explain 3-4 papers)

×