Information Systems Abstract ID#483
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Large Scale Hierarchical Text Classification Changing the way you search on the Internet
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
AGENDA • • • • • • • •
Abstract Introduction Goal Method Data Conclusion References Acknowledgments
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Abstract • With the amount of data available on the net, the million dollar question is how to extract valuable information from the humungous data available in the shortest time possible • More often than not, one does not know what exactly to search for and how to phrase it so that only desired results turn up. • LSHTC serves/tries to solve this problem.
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Introduction • Searching information online can be monotonous and frustrating • One needs to know how to search and what to search
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Goal • Improve the search experience • Reduce time spent on searching • Increase time spent on learning things that have been searched for • Correct the problem of having to spend about 2 hours a day finding information via search engines – correction… looking for information
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Method - Existing • 3 types – – Powered by robots (crawlers/ants/spiders) – Powered by human submissions – Combination of the two
• Crawler based engines visit website, read words and index it • This method is flawed
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Method - Existing
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Method - Proposed • Huge datasets created on topics thereby creating one of the largest database of databases known to mankind • Prototype – Plant database • Addition of data increases wealth of information
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Method - Proposed
Data
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Data
Data
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Conclusion • The proposed method can revolutionize the search mechanisms • Will reduce the time spent searching by more than 70% • What’s more you get more time off the computer… err.. can’t guarantee that
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
References • Data.gov - http://catalog.data.gov/dataset/gnispopulated-places9eda5 • USDA website http://plants.usda.gov/java/imageGallery?txtparm=&c ategory=sciname&familycategory=all&duration=all&gr owthhabit=all&nativestatus=all&wetland=all&stateSele ct=all&artist=all©right=all&imagetype=all&cite=all &location=all&viewsort=text&sort=sciname&submit2.x= 65&submit2.y=4 • Webopedia http://www.webopedia.com/DidYouKnow/Internet/Ho wWebSearchEnginesWork.asp • Go-gulf - http://www.go-gulf.com/blog/online-time/
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Acknowledgements • Chaiyaporn Mutsalklisana • My team – Adheesh Kulkarni, Gaurav Mahajan, Shreyas Katariya, Vashita Sharma
Large Scale Hierarchical Text Classification Changing the way you search on the Internet
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
AGENDA • • • • • • • •
Abstract Introduction Goal Method Data Conclusion References Acknowledgments
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Abstract • With the amount of data available on the net, the million dollar question is how to extract valuable information from the humungous data available in the shortest time possible • More often than not, one does not know what exactly to search for and how to phrase it so that only desired results turn up. • LSHTC serves/tries to solve this problem.
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Introduction • Searching information online can be monotonous and frustrating • One needs to know how to search and what to search
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Goal • Improve the search experience • Reduce time spent on searching • Increase time spent on learning things that have been searched for • Correct the problem of having to spend about 2 hours a day finding information via search engines – correction… looking for information
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Method - Existing • 3 types – – Powered by robots (crawlers/ants/spiders) – Powered by human submissions – Combination of the two
• Crawler based engines visit website, read words and index it • This method is flawed
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Method - Existing
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Method - Proposed • Huge datasets created on topics thereby creating one of the largest database of databases known to mankind • Prototype – Plant database • Addition of data increases wealth of information
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Method - Proposed
Data
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Data
Data
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Conclusion • The proposed method can revolutionize the search mechanisms • Will reduce the time spent searching by more than 70% • What’s more you get more time off the computer… err.. can’t guarantee that
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
References • Data.gov - http://catalog.data.gov/dataset/gnispopulated-places9eda5 • USDA website http://plants.usda.gov/java/imageGallery?txtparm=&c ategory=sciname&familycategory=all&duration=all&gr owthhabit=all&nativestatus=all&wetland=all&stateSele ct=all&artist=all©right=all&imagetype=all&cite=all &location=all&viewsort=text&sort=sciname&submit2.x= 65&submit2.y=4 • Webopedia http://www.webopedia.com/DidYouKnow/Internet/Ho wWebSearchEnginesWork.asp • Go-gulf - http://www.go-gulf.com/blog/online-time/
Graduate Category: Engineering and Technology Degree Level: Information Systems Abstract ID#483
Acknowledgements • Chaiyaporn Mutsalklisana • My team – Adheesh Kulkarni, Gaurav Mahajan, Shreyas Katariya, Vashita Sharma
