Q&A

So how do they do it?

1) They are collecting data (data curation), verifying it and making it computable. It’s partially computer and human edited.
2) Their algorithm and computations. Roughly 5 to 6 millions lines of mathematical code sorting out the data.
3) Linguistic analysis, they interpret the English language and map it into their search. They de-tangle the ambiguity of online search.
4) They automate the presentation for easier viewing. It picks out what is important. Again, it needs human intervention to make it work.

How do they keep the information current? e.g how would they keep current of the latest swine flu data?

There will soon be a mechanism for people to add data, it will then be processed partially by human and computer. So I guess it will work much like a wiki type social collaboration, although it will be more stable than a wiki as all the data will be edited before going live.

How many people will be needed to edit and bring in new data to keep Wolfram Alpha updated?

Wolfram was quoted in saying that he envisages 1000 people “shovelling” information into the WolframAlpha algorithm.

Where is the data retrieved from?

The semantic web is obviously an area to retrieve information but a large amount of it is retrieved from private and licensed databases like universities and governments.

Can it mix scientific information with popular culture i.e. music, celebrities etc?

There is a “shallow” amount of data available and is much simpler, for now it could probably measure the height difference of Britney Speares to that of Morgan Freeman but as Wolfram says, “Popular culture is not true data”. There are linguistic horrors which make it ambiguous i.e. because there are so many novels, movies, music etc, it could interfere with scientific data. However it could tell you box office sales for a particular movie.

Wolfram says that if you could imagine a library, they have covered roughly 90% of it, for now… It’s still growing…

Quoted from Twine

It doesn’t simply return documents that (might) contain the answers, like Google does, and it isn’t just a giant database of knowledge, like the Wikipedia. It doesn’t simply parse natural language and then use that to retrieve documents, like Powerset, for example.

Instead, Wolfram Alpha actually computes the answers to a wide range of questions — like questions that have factual answers such as “What is the location of Timbuktu?” or “How many protons are in a hydrogen atom?,” “What was the average rainfall in Boston last year?,” “What is the 307th digit of Pi?,” or “what would 80/20 vision look like?”

So what does this all mean for marketing your business online?

Well, you won’t have to mouth off your internet marketer for not changing your online marketing strategy just yet. Wolfram Alpha won’t affect your business now but looking into the future, it could.. What Stephen Wolfram has created is an algorithm that can punch out answers for you which quickens our research time. At the moment Google or any other search engine can’t offer this but we can see that it’s now possible. I’ve been toying around with the idea that rankings will soon not (if not already) be the best way to measure SEO performance but rather Analytics will be the best option. Wolfram’s algorithm strengthens my theory on this idea and I don’t think it’ll be long before Google moves this way as well. The major stumbling block though is that Wolfram Alpha is partially human edited and has taken a decade of development.

In my opinion I see Wolfram Alpha separating the internet into two divisions – Scientific/measurable data (Wolfram Alpha dominant) and entertainment/news (Google dominant).

I would like to write a follow up blog post to this once Wolfram Alpha has gone live and I have tested it for myself. If anyone has questions or comments please feel free to submit them below.