The research we do at Helioid involves a lot of reading. With some notes and summaries included, here is a list of the literature we’ve focused on from 2008 to 2009:
Machine Learning
G. Lebanon, Y. Mao, and J. Dillon. The Locally Weighted Bag of Words Framework for Document Representation. Journal of Machine Learning Research 8 (Oct):2405-2441, 2007.
This articles concerns a method for evaluating the fitness of content topic models and document topic models based on the dissonance between a set of documents and the set of documents generated by composing inverse functions and applying them to the original set of documents. A document generating function is applied to a topic generating function that is applied to the original set of documents. In order to compare topics, one can look at the original set of topics compared to set of topics generated by apply a topic generating function to the documents generated by applying a document generating function to the original set of topics.
Continue Reading...Google’s failure to present the user with relevant results stems from an inability to capture user feedback and implement a method for users to provide feedback. The failure in part stems from the assumption that a user wants to go through their search results in a linear fashion.
Here are a couple ways in which Helioid’s non-linear approach to searching helps solve this problem.