WordSeer - Reducing Your Time In Stacks

Certain topics, like literary passages and history require a lot of research to form a thesis. Technologies like Google Scholar have smoothed the process to a certain extent but one still has to swim through thousands of results before filtering out the needed few.

In response to this, UC Berkeley graduate Aditi Muralidharan has designed a sophisticated text-analyzing tool, WordSeer, as a part of her doctoral thesis. WordSeer aims to speed literary searches for humanities scholars and other researchers. Yes, there are many tools today that drive searches for words and phrases, but WordSeer packs the ability to distinguish grammatical structure as well as stylistic features. The program uses structure language to associate human logic and interpretation of search results.

#-Link-Snipped-#

 

For academic searches, the target is usually obscure rather than a specific keyword, hence a search accommodating each and every word in the sentence is needed. WordSeer puts forward an attempt to solve this by employing a technology called Natural Language Processing. Natural Language Processing makes the search broad using understanding about parts of speech, usage and word relationships. Though, Natural Language Processing as a concept is 20 years old, the execution was till now only possible using a complex programming code. With WordSeer, Aditi Muralidharan has provided a simple User Interface for this technology.

For example, a search asking “What does God do?” and “How was God described?” returned thousands of results and ranked the results by frequency of usage. The search revealed that attributes such as great, wise and merciful and actions such as bless, give and grant were in majority.



Source: #-Link-Snipped-# Image Credit: #-Link-Snipped-#