Personal View site logo
Google rank will soon depend on your match with trend
  • Vitaliy_KiselevVitaliy_Kiselev

    Google's search engine currently uses the number of incoming links to a web page as a proxy for quality, determining where it appears in search results. So pages that many other sites link to are ranked higher. This system has brought us the search engine as we know it today, but the downside is that websites full of misinformation can rise up the rankings, if enough people link to them.

    A Google research team is adapting that model to measure the trustworthiness of a page, rather than its reputation across the web. Instead of counting incoming links, the system – which is not yet live – counts the number of incorrect facts within a page. "A source that has few false facts is considered to be trustworthy," says the team ( http://arxiv.org/abs/1502.03519v1 ). The score they compute for each page is its Knowledge-Based Trust score.

    The software works by tapping into the Knowledge Vault, the vast store of facts that Google has pulled off the internet. Facts the web unanimously agrees on are considered a reasonable proxy for truth. Web pages that contain contradictory information are bumped down the rankings.

    Do not support main official POV that flood all high ranked web sites? Bad for you and your site.

    http://www.newscientist.com/article/mg22530102.600-google-wants-to-rank-websites-based-on-facts-not-links.html

  • 10 Replies sorted by
  • CrazyPeteCrazyPete

    1. Who holds the keys to this Knowledge Vault?

    2. How will this work for websites where the content is something other than a recitation of "facts" that can be found in many other places on the internet?

  • Vitaliy_KiselevVitaliy_Kiselev

    Who holds the keys to this Knowledge Vault?

    It is fully private Google thing. Except Google I am sure government and NSA will have their say.

    How will this work for websites where the content is something other than a recitation of "facts" that can be found in many other places on the internet?

    It is algorithms. It'll analyze text and if it founds hints that your point of view or some statements contradict to one set by KV it will reduce your search rank.

  • goannagoanna

    Follow the money trail ...

  • LOMLOM

    But isn't it similar to how it works generally. In academia peer review for publishing generally means that your views will be veted by those who represent the main stream view. It is also well know that as a PhD student your research dare not veer much from the wisdom of the herd that went before you. "Open Source" as a trend itself in publishing has to buck the entrenched KV and is generally still regarded as inferior.

    On the other side of the coin, ironically it is the private commercial publishers who do maintain a system that does strive to present the best research and as well maintain an accessable KV. In other words they are more trustable than the institutions themselves since the institutions have agendas that may be suspect (money based) because of adherance to govenrment funding objectives.

  • Vitaliy_KiselevVitaliy_Kiselev

    In academia peer review for publishing generally means that your views will be veted by those who represent the main stream view.

    Well, it is one of causes of current science state. And it is not good.

    But people are not algorithms, and especially not algorithms where you can pay to be excluded or included in KV.

    If they will implement it in full it'll be extremely powerful system to fight any unnecessary views.

  • LPowellLPowell

    From the abstract:

    We propose a way to distinguish errors made in the extraction process from factual errors in the web source per se, by using joint inference in a novel multi-layer probabilistic model.

    IOW, we will guesstimate whether your facts or our automated parse of your website are most likely rong.

  • goannagoanna

    While Google's profitability is impressive, they'll be watching out for their future revenue as well. Their current user-profiling model produces quite variable results according to the searcher; this commercial practice is not really compatible with correct results.

    Google's shareholders are risk-averse and don't necessarily agree with the company's wasting money on pseudo-altruistic activities. (Can we include factual search results here?)

    Google will be aware that its advertising model profits are not future proof; net users and clients can move quickly to whatever works for them. There are other ways to advertise and sell - just as there are some of us who have found other ways to search for quality information.

    See The next Google?

    http://ww.personal-view.com/talks/discussion/comment/190771#Comment_190771

  • goannagoanna

    FYI (and also a possible hint to Google needing to compete with other search engines who index other stuff besides websites...)

    This is old:

    For some time, Google has indexed orphaned web pages (for example, you may have your private calendar on a page which is kept private by not being linked to by any other web pages). Google used to do this by following users' "Google search bar" installed on their devices.

    But this is NEW:

    Google has stopped honouring the "robots.txt" directives not to crawl. So far, Google's robots are regularly crawling previously private - and protected web pages.*

    What will be next? Will our private "Birthday Greeting to my dear daughter" page appear in total strangers' web search results?

    • In going ahead and looking at our computers' settings, Google did give us all fair warning, after all - It's hidden in the terms of usage of the Google search bar. As for Google starting to disobey the time-honoured "robots.txt" request to "no-index" or "no-follow", Google have said for some time they prefer us to protect our web pages in Google's own way [using an .htaccess directive].
  • Vitaliy_KiselevVitaliy_Kiselev

    Google has stopped honouring the "robots.txt" directives not to crawl. So far, Google's robots are regularly crawling previously private - and protected web pages.*

    First, it is just recommendation to robots :-). Second, you need to check if they actually show this in search results.

  • Vitaliy_KiselevVitaliy_Kiselev

    Since 2013 Epstein and colleagues have conducted a number of experiments in the US and India to determine whether search results can impact people’s political opinions.

    Epstein points out that about 50 percent of our clicks go to the top two items on the first page of results, and more than 90 percent of our clicks go to the 10 items listed. And of course Google, which dominates the search business, decides which of the billions of web pages to include in our search results, and it decides how to rank them.

    But surely, Epstein thought, a top search result would have only a small impact on a person’s political choices. Not so! To Epstein’s surprise, in his initial experiment he found that the proportion of people favouring the (bogus, skewed) search engine’s top-ranked candidate increased by more than 48 percent! Also, 75 percent of the subjects in the study were completely unaware that they were viewing biased search rankings.

    He conducted several more experiments, including one that involved more than 2,000 people from all 50 US states. In that experiment, the shift in voting preferences induced by the researchers was 37 percent, and as high as 80 percent in some demographic groups.

    http://www.mauldineconomics.com/outsidethebox/the-new-mind-control

Add a Comment