trendpreneur

In amongst all this discussion of whether the web is dead, kicked off by the eponymous Wired article, it’s hard not to feel either depressed and defensive or jubilantly optimistic — depending on which pie you have fingers in.

Apps are great. You have control, you can do a lot more with native code, you only need to worry about a single launch platform (to some greater or lesser extent), you can easily make money, you can have an offline experience.

The web is great. You have control, you can iterate, you can experiment with pricing and revenue models, you can pull in elements of other services on the fly, you can share.

The argument is often phrased as a closed vs open fight, a standards vs haphazard war. Of course if you control a walled garden you can specify exactly what playing inside it should feel like. The key element is that it’s walled; if nobody else can come and play, you hinder your own progress as well as your competitors’.

That’s why I was excited to see the Lifehacker Android Pack. Apps? Yup. Closed? Sure. But the very fact that an app broke out of the garden to become a secondary marketplace – to fix all that’s wrong with the primary marketplace – and that it’s possible to do things like recommend a ‘pack’ of apps through what is, itself, an app — that’s pretty cool.

We’ve heard before that plugging holes is not sustainable; of course not. Building on top of another service and hoping they don’t fix the problem you’re solving? Very tricky if you’re a startup looking to take millions of other people’s money. Absolutely fine if you’re an experiment that’s patching up a serious flaw until the powers that be notice you and offer to buy your duct-tape from you.

And apps like AppBrain, ironically, embody what I love about the web. Got data? Got an API? Mix it up and republish it and add your own value on top, then let the world enjoy it. Most of the recent few years’ shift in landscape and thinking wouldn’t have happened without the concept of an API suddenly becoming hot stuff, and the concept of a freely-accessible API being deeply ingrained alongside. But it makes so much sense. Why limit yourself to what your own company’s brains can think up, when people the world over are begging to do stuff with your data?

Anyway, this is why I’m excited by the Chrome web store, as it has the potential to marry up a lot of these concepts – though maybe not initially. But it’s the thought and the attitude that counts, right?

Are you in your late twenties or early thirties? Did you grow up in Britain and have fond memories of Britpop, Blair and Big Brother? (Scratch that last one..)

I’m working on a wee project combining video and music of the times, and I’d love to collect memories from the time. Wow, that makes it sound like history… Anyway, we all have fond flashbacks to the age of heat reactive tie-dye shirts, and although my middle-of-the-road upbringing is chock full of happy reminiscences, I want more variety!

So, if you have a story, a memory, a moment to share from around 1992 to 2000 – for me, my high school years – and especially if it’s closely related to (or described by) a song popular in the same time period – please share it with me :) You can comment here/on Facebook, or email me (mail at jennielees dot net). Oh, and it doesn’t have to be meaningful or funny – just true.

Hmm, should I start a site to collect these or keep them private?

Here’s mine: our endless school assembly pop culture parodies. I remember dressing up as Mikey from Boyzone (I secretly wanted to be Ronan), doing a Grease ripoff for a departing Latin teacher that was, in retrospect, dangerously suggestive, and for some reason I posed as Renton from Trainspotting in front of our class blackboard in N3. God knows why. Anyway, cue montage! I wonder if kids these days do the same type of thing…

At Twitter’s Chirp conference today, the company announced an interesting move. Currently, you can attach a location to your tweets, and not just co-ordinates either; you can boast your neighbourhood and city.

The logical extension, which Twitter will roll out this quarter, is attaching places to tweets. Hmm, sounds somewhat familiar…

Thing is, this leads to an interesting gap. Instead of check-in fatigue, this could reduce the need to check in at venues; send a tweet, and it gets you automatically checked in at that venue, maybe even posting your tweet as a tip for that place.

One issue is the back-channel that occurs when you check in using a service like Foursquare. It’s good to get points and badges and shiny things; if check-ins are automatic, you don’t get any of that.

Still, it’s a nice concept for someone to implement, one day. Someone who isn’t me.

Finally, a post topic that combines both sentiment analysis and the meta-world of professional blogging!

I usually like TechCrunch for the most part, but these two articles have annoyed me: ‘Sentiment is split on the iPad‘ and ‘More iPad Sentiment Analysis‘. Both use poor, crude methods of sentiment analysis to produce posts full of fluff and pretty graphs. Result? Whatever point the blogger wanted to make. (You know what they say about statistics).

A quick rundown of the problems: Spurious classification algorithms, poor data sizes, and non-credible results. An algorithm that analyses every piece of traffic on Twitter and comes up with “51% positive, 49% negative” is Just Plain Wrong. There’s going to be a ton of stuff in the middle, unclassifiable, undecided, even just retweets of blog posts with the word in the title, and any graph should reflect that as well. Stripping out the neutral, a result of 51/49 just seems completely nonsensical to me, and I’ve been working with Twitter sentiment for a long time now.

It’d also be very interesting to know what methods the classifiers use, probably available with some digging, but I fear it’s manual keyword lists that some poor sod had to draw up — “hmm, I think if someone says the iPad is ’stupid’ that’s probably negative, yah?”.

Attensity does better, but what on earth does “not thrilled” mean (weak negative?) and again, where’s the neutral or noise aspect? It’s valuable to know just how many tweets were about the iPad, and how many of those were about sentiment. What if a TechCrunch headline with a negative word got retweeted 2000 times? That’s what we in the trade call “skew”. Plus, classifying on a small sample is just crazy. Why? Surely it can’t be computational limits; were these the only tweets with sentiment information? That’s useful data! Why throw it all away…

It also looks like there are some great leaps in logic in terms of distinguishing between “Like the iPad because it might replace iPhone” and “Don’t like the iPad because it won’t replace my iPhone”. How do you automatically extract the difference between “Can’t replace battery” and praise for the battery life? Sigh.

Plus, there’s the key mistake of not showing error, accuracy bounds, or mistakes. Both posts assume the algorithms are 100% correct. While that makes for some pretty graphs, it just isn’t true, and with no idea of sample size or result size (e.g. for the battery category above) then a result of 5% could just mean one out of a total of twenty tweets with the word battery in was negative. It’s the same for intent to purchase. Not every tweet will have any kind of intent, so if you just took the tweets containing “will” “buy” “iPad” or “won’t” “buy” “iPad”,

Of course, the reason I’m most annoyed at these posts is that I could have helped put together a custom dataset and classifier to provide much more detailed data, and didn’t. But while I can’t go back in time and change things, I can at least point out the flaws in using off the shelf graphs to meet your daily post quota as a pro-blogger.

Disclaimer: This post is personal opinion, the views expressed here are not those of Google, and not influenced by any relationships the poster may have with the Big G.

There have been arguments raging on and offline about paywalls, the commons, old media versus new media, and ‘information should be free’ for — well, it feels like forever now. One of the (many) components of new media under fire is the army of filthy idea-stealin’ bloggers, people who merrily subscribe to paid content and then go and paraphrase it on their free-to-view blogs (or in some cases, just copy it). Paul Carr makes an excellent point about the commoditisation of facts, the human need for information and thus the Internet hivemind’s tendency to trend towards free.

Information being free is good, for obvious reasons, unless you’re someone who wants to get paid to create it. There are plenty of arguments for well-crafted columns, investigative journalism, paid political pundits and so forth. But here’s a thought about the oft-maligned practice of reblogging, rephrasing, and retweeting.

Language is variable.

The more ways an idea or piece of information is expressed linguistically, the easier it is to find — it’ll match far more search queries, as a simple starting point. Although, in an echo of the Sapir-Whorf hypothesis, perhaps expressing an idea in multiple languages, or with different phrasings and words, could change the way people think about the idea. Even if this happens, the idea reaches far more people than it would have if it were confined to one site, in one language, by one author.

From Google’s point of view, if someone takes a New York Times article, paraphrases it, and links back to it, the data miners jump for joy. Beautiful, delicious data. We learn new things about the relationships between words and concepts — maybe one article said climate change but another global warming. The link-back gives us contextual data that can help too. (Linking to a climate change article with the text “This article on global warming”, for example).

Of course, paraphrasing and rewriting has been going on for years, a staple of the essay or lit review. But as with voice recognition, having the power to implement and use a feedback loop at world-scale is a mind-blowing thing. Google has the power to build an entire semantic web out of paraphrased blog posts, and that’s before we even look at contextual links in Wikipedia or Twitter link summaries. If that’s scary, just think of the magic that happens when you search for something and get a result that isn’t the exact terms you entered, but is the exact concept. With a bit of data, intelligence and an army of semantic web PhDs, it just could happen.