Tuesday, 21 February 2012

Article Summarization using Machine Learning

There's nothing more fun than writing a machine learning algorithm while watching Terminator. I decided to take a crack at summarizing tech articles without any knowledge of existing algorithms. I used ruby to make a script that stored memory in one file, and filtered output in another file. The filtered output is somewhat "choppy", but I think it did an alright job. The links to the articles, as well as my summaries can be found below.

Facebook profiles accurately predict job performance

profile That s the advice conscientiousness extraversion agreeableness and neuroticism Six months later the researchers compared evaluations of are of course subjective but job seekers shouldn t necessarily worry that they need to clean up their Facebook profile For instance Don Kluemper one of signaled openness to new experiences and adventurousness while also looked at evaluations drawn from Facebook profiles and academic success Kluemper says that report which included a larger sample 244 and found academic Kluemper Despite the studies companies shouldn been enough research yet to definitive connection between profiles and job offers indicator done

Google to sell heads up display by years end

data connection tilting scroll month We are told it is very quick learn and once user adept at navigation it becomes second nature almost indistinguishable low resolution built actively discussing and dozens futuristic projects One most rather than like those that people see on smartphones The glasses will send data to looked at about potential business models with Disruptions column last year Apple engineers are also exploring wearable computing but the company is taking a different route focusing on computers that strap around someone s wrist Last week The San Jose Mercury News discovered plans by Google to build a 120 million electronics testing facility that will be involved in testing precision optical technology


other just watch video above It s raises some very valid points Change is hard for everyone Switching from Office to Google Docs isn t an easy switch Many features are still missing from Google s productivity suite It s not a direct replacement Typing in the cloud editing Microsoft Office might be disappearing from some shops but others rely on it as much as the office s coffeepot Google will no doubt fire back with its own quirky advert You can always count on Microsoft and Google s marketing department for a good show

It obviously needs some tweaking, but I think it's a start for an hour's work.

No comments:

Post a Comment