Analysis, thoughts on Google Introducing the Knowledge Graph: things, not strings (Updated)

Yesterday Google (NASDAQ: GOOG)  announced on its Blog, a new relational search feature to be gradually rolled out over the coming days. Whilst formally not new, it was considered novel enough to make it official, in likely full awareness of the ensuing media fuzz.

2500 Mentions after a 16h window (Data provided by Topsy Inc)


This announcement involved several significant points and discoveries for me:


  1. Google overtly cleans-up its code, often including state of the art technologies present in the Web. From that perspective the lack of a keyword tag <meta keywords="..." />  is virtually a statement in itself to SEO's, SEO related businesses and website administrators: Meta-keywords are obsolete! 
    Meta-tag keywords have the tremendous drawback that they rely on the truthfulness of the content-provider, rather than relying on information which is generated by own algorithms on the nature of the content.In other words, why involve the third party stating the truth if you can just mine the truth, at the expense of some CPU cycles. Trying to beat the system thus at least involves algorithms thus some effort, as opposed to none
  2. When looking at Google blog's content traffic, the use of image sprites to avoid HTTP roundtips are an ubiquitous asset in Google's web repertoire for 3+ years. Interestingly some of the sites JavaScripts also passed along base64 encoded images. Moreover, Chrome can directly read such base64-strings with a recognized MIME-type attribute from the url-address bar, and a drag and drop operation will bring these back into binary form eating up space on your hard-drive (or... cloud-drive?) You can try to copy and paste the content below into your address bar. It will show an animated loader-gif image. Update: Works in Opera >10.5, Firefox as well (not all MIME-types).
    data:image/gif;base64,R0lGODlhIAAIAKECAEVojoSctMHN2QAAACH/C05FVFNDQVBFMi4wAw
    EAAAAh+QQFCgADACwAAAAAIAAIAAACFZyPqcvtD6KMr445LcRUN9554kiSBQAh+QQFCgADACwCAA
    IAEgAEAAACD4xvM8DNiJRz8Mj5ari4AAAh+QQFCgADACwCAAIAHAAEAAACGJRvM8HNCqKMCCnn4
    JT1XPwMG9cJH6iNBQAh+QQFCgADACwMAAIAEgAEAAACD5RvM8HNiJRz8Mj5qri4AAAh+QQFCgAD
    ACwWAAIACAAEAAACBZSPqYsFACH5BAUUAAMALAAAAAAgAAgAAAIOnI+py+0Po5y02ouzPgUAOw==
  3. The cat is out of the bag: Google is indeed heavily invested in Graph databases  and Process graphs (Pregel: a system for large-scale graph processing, Greg Malewicz, Google, Inc.; Matthew Austern, Google, Inc.; Aart Bik, Google, Inc.; James Dehnert, Google, Inc.; Ilan Horn, Google, Inc.; Naty Leiser, Google, Inc.; Grzegorz Czajkowski, Google, Inc., SIGMOD 2010, Published by ACM 2010). Owing to their algorithmic nature graph databases are well amenable to parallelilzation particularly the Map-Reduce computing paradigm (Parallel Data Processing with MapReduce: A Survey - SIGMOD 2011 by KH Lee - PDF). 
  4. Finally, it is becoming clear that Google cannot afford to make any major product without offerings towards the/'its' mobile market. With Android now having a significant lead over iOS, that strategy is a no-brainer. Now I am gradually realizing it as well, manifested in recent dabblings in Phonegap
    Update: The mobile connection of the knowledge-graph has been more or less affirmed through this official blog post.
    Indeed targeting the mobile market is most sensible, due to the inherent limitations of the input device imposed by mobile form-factors. 'Power-googlers' (i.e. users that spawn a new tab every minute and enter a new search every ten seconds) are likely to remain unimpressed by the new additions. 
  5. (Google is increasingly no longer content with half of the search-site remaining blank? I thought that 'is Google'. Is the increasing screen-real estate at fault or is the lack of competitors leading to general discontent ;) ?)
LihatTutupKomentar