Please use this identifier to cite or link to this item:
http://hdl.handle.net/10603/270625
Title: | Computational Immune System for Ambiguity Reduction in Human Computer Speech Interaction |
Researcher: | Ghosh Papri |
Guide(s): | Chingtham, Tejbanta Singh and Ghose, Mrinal Kanti |
Keywords: | Human Computer Speech Interaction |
University: | Sikkim Manipal University |
Completed Date: | 2019 |
Abstract: | The advancement in interactive technology since last few decades has evolved the newlineHuman-Computer Interaction as a major player in the areas of artificial intelligence newlinecognitive science psychology sociology anthropology, management science and newlineengineering. Most of the human-machine interactive applications today use speech as newlinean input. Over the years, speech recognition has become the highly efficient tool with newlineminimal error. In sentence level speech recognition, the system identifies correct newlinehomophone word through lexical analysis; but in word level speech recognition, newlinehomophone is a major ambiguity that minimizes the system efficiency. The existing newlinespeech recognition system uses outsized system memory and the homophone newlineambiguity still exists in word level recognition technique.The Artificial Immune System is an emerging area of Artificial Intelligence.However the performance of Artificial Immune System has not been extensively tested on Human Computer Interaction.In the present study, an attempt has been made to explore the optimization and newlinereduction of ambiguity due to homophones at word level speech recognition. Two newlineapproaches have been proposed and developed to optimize the space complexity in newlineSpeech Recognition Technology and to reduce homophone ambiguity from Word Level Speech Recognition. The first algorithm uses the system inbuilt dictionary to newlinereduce space complexity of Speech Recognition System by obtaining the synonyms.The second approach involves recognizing all the homophones from a same homophone set in the Word Level Speech Recognition (WLSR). An Artificial Immune System with Affinity Maturation Technique is configured to recognize homophones based on their priorities. A self-maturation technique which is inspired from Artificial Immune System is applied to modify or update the homophone newlinedatabase based on the frequency of usage of a homophone. The system and method is configured to recognize two or more than two homophones from same homophone set by updating the priorities of homophones |
URI: | http://hdl.handle.net/10603/270625 |
Appears in Departments: | Department of Computer Science and Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
01_title.pdf | 446.35 kB | Adobe PDF | View/Open | |
02_certificate.pdf | 1.32 MB | Adobe PDF | View/Open | |
03_abstract.pdf | 346.68 kB | Adobe PDF | View/Open | |
04_contents.pdf | 804.72 kB | Adobe PDF | View/Open | |
05_chapter 1.pdf | 797.57 kB | Adobe PDF | View/Open | |
06_chapter 2.pdf | 646.59 kB | Adobe PDF | View/Open | |
07_chapter 3.pdf | 1.06 MB | Adobe PDF | View/Open | |
08_chapter 4.pdf | 1.33 MB | Adobe PDF | View/Open | |
09_chapter 5.pdf | 1.34 MB | Adobe PDF | View/Open | |
10_chapter 6.pdf | Attached File | 872.37 kB | Adobe PDF | View/Open |
11_chapter 7.pdf | 585.04 kB | Adobe PDF | View/Open | |
12 _references.pdf | 489.03 kB | Adobe PDF | View/Open | |
13_appendices.pdf | 3.11 MB | Adobe PDF | View/Open |
Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).
Altmetric Badge: