Posts Tagged ‘algorithm’

Give Em This…

Friday, September 23rd, 2011

Look no further!

TIN IT OR BIN IT is your weekly social media need-to-know.

Hot off the press we deliver the week’s most ground breaking social media news,
oh-no-you-didn’t stories, tips you can’t afford to miss or simply our opinion on trends in the digital landscape.

Lastly, we’ll tell you whether it’s tin material or bin material!

View them all here on our Facebook page.

———————————————————————————————————————–———–

Tin It Or Bin It – 23 September 2011 – Give Em This…

Sometimes finding the perfect gift for your nearest and dearest is, if we are being honest, a bit of a struggle. If you are stuck for ideas, then GiveEmThis.com might just be what you are looking for.

This free web app, by Imply Labs, is a personalised gift engine that algorithmically analyses your friend’s social media updates to recommend relevant gift ideas that your Facebook friends actually want.

The system connects to your social media and analyses your friend’s interests based on their likes, dislikes, interests and other data stored in their profile. GiveEmThis.com then recommends original gift ideas relevant to your friend’s interests and activity.

GiveEmThis.com doesn’t select one perfect gift (that’s up to you) but looks at countless options to show you the things they will love the most.

The app works by calculating your Facebook friend’s online data, behaviour and patterns. Events, status updates, shared links and “Likes” are all included in the algorithmic calculations.

It doesn’t just evaluate Facebook data. Twitter is also analysed, although the downside it that it isn’t as detailed.

The programme is still in its early stages, with only Amazon being used to generate gift ideas. However there are plans to expand, with companies potentially being able to add their products to the system, and Google+ and LinkedIn integration.

Now, don’t get us wrong, we really like the practical application of this gift generator…so we should TIN IT, but we are going to BIN IT purely because we don’t like Facebook and Twitter knowing so much about us. And Christmas just wouldn’t be the same without those weird and wonderful random presents we get but don’t want!

This gift generator is a BIN IT.

Give Em This

Man versus machine

Monday, September 5th, 2011

Look no further!

TIN IT OR BIN IT is your weekly social media need-to-know.

Hot off the press we deliver the week’s most ground breaking social media news,
oh-no-you-didn’t stories, tips you can’t afford to miss or simply our opinion on trends in the digital landscape.

Lastly, we’ll tell you whether it’s tin material or bin material!

View them all here on our Facebook page.

———————————————————————————————————————–———–

Tin It Or Bin It – 2 September 2011 – Man versus machine

For all the literary lovers out there, there is now a website that can recommend you books using an algorithm. Launched last week by Aaron Stanton, BookLamp is to books what Pandora is to music.

The website analyses a book’s style of writing and assigns numerical values to the ‘Story DNA’, which is described as the breakdown of the settings, themes and characters.

We don’t know about you, but we choose books through human recommendations – from friends, family, colleagues, book reviewers in newspapers and Amazon user reviews.

Obtaining electronic recommendations from an algorithm that has not read and experienced the book just seems a bit dubious. We love technology but sometimes you just need a human touch.

BookLamp can’t distinguish between good prose and sloppy writing styles, so it is lacking the quality control that you usually get from someone you know and trust.

Also, the BookLamp engine can sometimes recommend a ‘zinger’ – which is a book that makes sense from an algorithmic perspective but not from a human view. This seems to be a major problem and something they will need to work on to perfect.

So far there are only 20,000 titles in their database, which means that your choices are fairly limited at this stage until more publishers come on board.

BookLamp is a work in progress. It sounds like a good idea initially, but on closer inspection there are some flaws.

We are not entirely convinced. Think we will just stick to old-fashioned book recommendations from friends and family for now, and come back to this once the BookLamp engine has had time to expand and fine-tune itself.

BIN IT…for now.


Sign up for Company Subscribers
* = required field