SiliconFilter

Don’t Want Google to “Correct” Your Searches? You Can Now Search Verbatim

/

Until earlier this year, you could do a Google search and use the ‘+’ operator in front of any word to make sure that Google would search for this specific term. Now that Google is moving towards using ‘+’ as a way to find Google+ profiles, though, this option is gone. Instead, Google asked users to use double quotes to ensure that none of Google’s usual corrections, personalizations or other changes are applied to this term. Now, however, after some vocal opposition against the disappearance of the ‘+’ search operator, Google is introducing a new tool that brings some of this functionality back: verbatim search.

verbatim_search_find_itThis new tool will be rolling out to all Google users over the next few days. Once it’s available for you, it will hide in the side bar, though, where you will have to click on “more search tools” and then look for it at the bottom of the list.

While few users are likely to really need this tool, it’s nice to have it as an option, though just adding quotation marks around a term seems to be a bit easier than clicking through to the “more search tools” option.

Algorithm Changes Coming, Too

Google also acknowledged that it plans to make some changes to when its “query broadening search improvements” triggers, which seems to be an admission that its current algorithms sometimes go overboard in trying to correct or broaden search queries.



6:46 pm


Wajam Wants to Make Your Social Search More Social

/

Social search is, without doubt, one of the hottest topics in the search engine business today. Google and Microsoft have made it the central focus of their latest search engine features and numerous small players are also trying to get a foothold in this nascent business. Among these smaller players is Wajam, a Canadian startup that lets you easily add social search results to virtually all of the majorsearch engines and shopping sites you use today, including Google, Bing, Amazon, Tripadvisor, Wikipedia, and Yelp.

The idea behind social search has always been intriguing, as there is, after all, a good chance that the links your friends share online are more relevant to you than other links. To make this really work, though, a social search engine needs to be able to easily tap into all your social networks, not just either Twitter or Facebook. That’s where Wajam shines. It lets you connect to all your favorite social networks and then indexes all the links (and the content of the pages these links point to) that your friends have shared. Then, when you search, it transparently pins these results at the top of your regular search results on your favorite search engine.

Among the nifty features here are the ability to also add your Google+ account and search through it – something that Google still doesn’t let its users do. You can also filter results so you just see photos or just the links a specific person has shared. Earlier this month, Wajam also added a location feature, which lets you easily see who of your friends live in a given city and what places your friends have liked there.

Earlier this week, I talked to the company’s founder and CEO Martin-Luc Archambault. According to Archambault, his team mostly consists of engineers, as the company runs its own servers and has to not just pull in a very large amount of data (my friends, for example, have shared more than 3.5 million links) but also rank it. The ranking, indeed, could still use some tweaking, but in general, the search results are relevant, though the best ones are often under the fold (by default, Wajam only shows one result).

Overall, though, Wajam has turned out to be quite a useful addition to my search arsenal, especially because it pulls in data from such a wide variety of sources.

wajam_wikipedia

Enhanced by Zemanta


4:12 pm


Better Search Results for All: Google’s Panda Update Goes Global

/

In its efforts to preserve the quality of its search results, Google rolled out the so-called Panda and Panda 2.0 update to its algorithm for searches in English earlier this year. Until now, however, these changes didn’t impact searchers outside of the English-speaking world. That’s changing today, however. Earlier this morning, Google announced that it has now brought its “algorithmic search improvements” to all other languages, with the exception of Chines, Japanese and Korean.

Impact: 6-9% of All Searches

According to Google, these changes will impact about 6-9% of all queries to the degree that users will notice the difference. The earlier Panda update for English queries was decidedly more aggressive, as it affected a good 12% of all searches.

While Google doesn’t explicitly say so, the originally Panda update was – for a large part – motivated by the proliferation of content farms that pollute search results with low-quality content written by badly paid freelancers. Indeed, companies like Demand Media were strongly affected by this change and lost a good amount of traffic because of it.

The content spam problem isn’t quite as bad in the rest of the world. It looks like Google clearly felt that the Panda update improved search algorithm worked well enough in other languages as well to roll it out globally.

Image credit: Flickr user Stéfan



4:07 pm


Google Autocomplete Gets Enhanced Predictive Powers

/

Google just announced an update to its autocomplete feature, which speeds up the search process by showing predicted searches while you type. Until now, Google mostly based its predictions by looking at the most popular searches. The problem with this, Google points out, is that the majority of search queries have never been typed in before and hence didn’t show any predictions. Now, however, Google is expanding this feature by “improving the predictive powers of autocomplete” for these seldom used queries as well by just looking at the last part of the query.

By just focusing on the last part of the query, this is obviously not as helpful as autocompleting longer queries, but will likely save users some time in the long run.

Google is rolling this new feature out today on google.com in English and will expand it to other languages and locales over time.



1:26 pm