I retired from personal blogging in July 2008.
But you can find me over at http://blog.xero.com.
That’s a big number ….
Quantifying Microsoft’s biggest purchase ever
“Who would have guessed years ago that this sort of ‘vertical consolidation’ would occur between the search/portal players and the ad networks,”
blog
Big companies are rushing to gobble up online advertising agencies. This only shows that the competition is fierce in web-technology domain and that is a good thing because new technology innovations will accelerate. From this morning’s Herald article, it quoted from the Youtube founders that Google is working on a new universal search engine that is able to combine search for text, video, image, maps, etc. The reason is to drive up earnings from advertising. The types of algorithms that are needed for such universal search engine are not new, although there will be improvement over existing ones. There are tons of research publications available on different techniques that could be used for such search engine, however I don’t believe that only Google is the only one developing it. The search algorithms that exist for multimedia type data such as image, audio, etc, are already available, but I think that Google and others are working on combining them (PageRank from link-search, Latent Semantic Indexing from text content, computer vision from image recognition, audio recognition (such as search for songs with similar tunes by humming the song) from digital signal processing, techniques.
There was a workshop at Stanford in 2006 on Algorithms for Modern Massive Datasets, where reps from Google, Yahoo, Microsoft and others attended. Some of of the algorithms (new & improvement of old ones) in the invited presentations are of interest to the type of universal search engine that Google is currently developing. The presentations slides and notes are available for free download from the site. So, I think that Google, Microsoft, Yahoo are working independently in developing their own universal search engine.
At least they didn’t buy yahoo (yet?).
I think MSFT is in danger of swapping being the lead player in desktop/server enabling technology for a role as a second tier player in online. Having a low cost outlier presence in peripheral businesses is good to help spot incoming competition. Refocussing the firm into playing catchup in a lower value business isn’t.
I’d be a lot happier to see them put $6bln into R&D to try and kickstart pattern recognition based technologies.
Microsoft has already been ahead by a wide margin in comparison to Google in the development of pattern recognition based technology. Just take a quick look at the topics in pattern recognition that they have published in Journals from ACM or IEEE, here:
“Web Search & Mining Group”
http://research.microsoft.com/wsm/
Note that the listed papers on the Selected Publications section only covers the year 2004. I have seen more peer review publications from Microsoft Research Center that had appeared in a varieties of ACM & IEEE journals for 2005, 2006 and recently upto February of this year , 2007. Google & Yahoo don’t publish their research work, while Microsoft does. I believe that the reason Microsoft do publish their research is because, most of the papers that I have come across are co-authored with researchers from different universities around the world. So, those academics who co-authored research papers with Microsoft team from their research center do want their work to be submitted or being published in peer review publication, since that is what academic live for, to able to publish their research work. I think that Microsoft then further develope commercial software applications based on those collaborative research work & publications with those outside academics. So, it is this further commercial development is what the public don’t know as they want to keep their IP to themselves.
I for one, do not believe that Google will be dominant forever, I still believe that Microsoft will overtake one day, and this is based on what I have read from their (Microsoft Research Center) research peer review papers that are available in the public domain. As for the Google universal search engine, one only has to look at the link shown above and see the topics that Microsoft had been doing research on for some years now, which are (algorithms) something just new to Google, and it could take Google a while to perfect those techniques. We only hear about Google building its own universal search engine, but we don’t know if Yahoo or Microsoft does. I had seen a Microsoft paper in 2003, that outline the concept of such search engine and it would be naive for anyone to think that the Giant is not aware of it or doing nothing about it.