SmartSearch: A Voice Sensing Personalized Mobile Web Search Application - PowerPoint PPT Presentation

Loading...

PPT – SmartSearch: A Voice Sensing Personalized Mobile Web Search Application PowerPoint presentation | free to download - id: 5a8a2b-NzQwY



Loading


The Adobe Flash plugin is needed to view this content

Get the plugin now

View by Category
About This Presentation
Title:

SmartSearch: A Voice Sensing Personalized Mobile Web Search Application

Description:

Title: Parallel Data Mining Optimal Algorithm of Virtual Cluster Author: Owner Last modified by: Amit Created Date: 4/9/2010 2:26:08 AM Document presentation format – PowerPoint PPT presentation

Number of Views:55
Avg rating:3.0/5.0
Slides: 13
Provided by: Own2293
Learn more at: http://impact.asu.edu
Category:

less

Write a Comment
User Comments (0)
Transcript and Presenter's Notes

Title: SmartSearch: A Voice Sensing Personalized Mobile Web Search Application


1
SmartSearch A Voice Sensing Personalized
Mobile Web Search Application
  • Team
  • Abilash Bhanoori,
  • Amit Chaube, Sumit Shrivastava

Date-12/06/2011
2
The Problem
  • Internet has huge amount of data.
  • Each user has different interest.
  • Each user has different interpretation.
  • Maintaining a session is difficult on desktop.
  • The small form factors of the mobile devices.

Personalized Web Search with Location Preferences
(ICDE 2010)
3
Significance to Mobile Computing
  • Mobile phones based Web
  • Search Application.
  • Usage convenience provided
  • by voice enabled input.
  • Context Awareness provided by
  • personalization of results.
  • Extracting and Maintaining
  • Session becomes easy.
  • Project Demonstration

4
SmartSearch Approach
  • SmartSearch is a convenient and
  • efficient way to search the web
  • using mobile phones.
  • Supports two input modes,
  • i.e. voice and keypad input.
  • Maintains User Profile.
  • Its a meta search engine which
  • uses results from Google, Yahoo,
  • and Bing (Yahoo! for now).
  • Personalizes the results using
  • the clickthrough history, and
  • users preference profile.

5
Voice Input
  • Using Googles voice
  • recognition API.
  • User still can edit the query
  • if not satisfied by the match.
  • Keypad input still available.

6
Result Retrieval
  • Yahoo! Boss API is being
  • used for getting the results
  • from web.

7
Result Retrieval
  • from web.

8
User Profiling
  • When a link is clicked
  • most important keywords
  • from the URL are retrieved
  • and stored.
  • Same query on the next
  • search will display different
  • results.
  • Similar Pages related to
  • what user clicked are shown.

Personalized Web Search with Location Preferences
(ICDE 2010)
9
Some Tech Stuff
  • Keyword extraction is done using tf (term
    frequency) idf (inverse document frequency).
  • Classification of query and URLs is done using
    Naïve Bayes Classifier.
  • Users Profile is stored in a file on the phone.

10
Conclusions and Future work
  • A very important and necessary approach to
    personalize results.
  • Its better to have our own Search engine and
    Speech Recognition.
  • Rigorous Testing of the application on multiple
    phones will result in better classification.
  • If pursued, it can prove to be a great commercial
    product.

11
Questions?
12
  • Thank You!
About PowerShow.com