Search using your terms, verbatim

Behind the simplicity of Google search is a complex set of algorithms that expands and improves the query you’ve typed to find the best results. Automatic spelling correction ([vynal] to “vinyl”) and substituting synonyms (matching [pictures] to “photos”) are just two examples of the improvements we make.

In most cases, Google’s algorithms make things better for our users - but in some rare cases, we don’t find what you were looking for. In the past, we provided users with the “+” operator to help you search for specific terms. However, we found that users typed the “+” operator in less than half a percent of all searches, and two thirds of the time, it was used incorrectly. A couple of weeks ago we removed the “+” operator, encouraging the use of the double quotes, which are more likely to be used correctly.

Since then, we’ve received a lot of requests for a more deliberate way to tell Google to search using your exact terms. We’ve been listening, and starting today you’ll be able to do just that through verbatim search. With the verbatim tool on, we’ll use the literal words you entered without making normal improvements such as

• making automatic spelling corrections
• personalizing your search by using information such as sites you’ve visited before
• including synonyms of your search terms (matching “car” when you search [automotive])
• finding results that match similar terms to those in your query (finding results related to “floral delivery” when you search [flower shops])
• searching for words with the same stem like “running” when you’ve typed [run]
• making some of your terms optional, like “circa” in [the scarecrow circa 1963]

You can access the verbatim search tool under “More search tools” on the left-hand side.

In addition to verbatim search, which will be rolling out to all users over the next few days, we’re also applying similar ideas directly to our algorithms, such as tuning the accuracy of when our query broadening search improvements trigger. In the meantime, if you want to search for a very specific term, be that [carosel] or the [etymology of sissors], give the verbatim tool a try.

Posted by Corin Anderson, Principal Engineer, Search

Ten recent algorithm changes

Today we’re continuing our long-standing series of blog posts to share the methodology and process behind our search ranking, evaluation and algorithmic changes. This summer we published a video that gives a glimpse into our overall process, and today we want to give you a flavor of specific algorithm changes by publishing a highlight list of many of the improvements we’ve made over the past couple weeks.

We’ve published hundreds of blog posts about search over the years on this blog, our Official Google Blog, and even on my personal blog. But we’re always looking for ways to give you even deeper insight into the over 500 changes we make to search in a given year. In that spirit, here’s a list of ten improvements from the past couple weeks:

• Cross-language information retrieval updates: For queries in languages where limited web content is available (Afrikaans, Malay, Slovak, Swahili, Hindi, Norwegian, Serbian, Catalan, Maltese, Macedonian, Albanian, Slovenian, Welsh, Icelandic), we will now translate relevant English web pages and display the translated titles directly below the English titles in the search results. This feature was available previously in Korean, but only at the bottom of the page. Clicking on the translated titles will take you to pages translated from English into the query language.
• Snippets with more page content and less header/menu content: This change helps us choose more relevant text to use in snippets. As we improve our understanding of web page structure, we are now more likely to pick text from the actual page content, and less likely to use text that is part of a header or menu.
• Better page titles in search results by de-duplicating boilerplate anchors: We look at a number of signals when generating a page’s title. One signal is the anchor text in links pointing to the page. We found that boilerplate links with duplicated anchor text are not as relevant, so we are putting less emphasis on these. The result is more relevant titles that are specific to the page’s content.
• Length-based autocomplete predictions in Russian: This improvement reduces the number of long, sometimes arbitrary query predictions in Russian. We will not make predictions that are very long in comparison either to the partial query or to the other predictions for that partial query. This is already our practice in English.
• Extending application rich snippets: We recently announced rich snippets for applications. This enables people who are searching for software applications to see details, like cost and user reviews, within their search results. This change extends the coverage of application rich snippets, so they will be available more often.
• Retiring a signal in Image search: As the web evolves, we often revisit signals that we launched in the past that no longer appear to have a significant impact. In this case, we decided to retire a signal in Image Search related to images that had references from multiple documents on the web.
• Fresher, more recent results: As we announced just over a week ago, we’ve made a significant improvement to how we rank fresh content. This change impacts roughly 35 percent of total searches (around 6-10% of search results to a noticeable degree) and better determines the appropriate level of freshness for a given query.
• Refining official page detection: We try hard to give our users the most relevant and authoritative results. With this change, we adjusted how we attempt to determine which pages are official. This will tend to rank official websites even higher in our ranking.
• Improvements to date-restricted queries: We changed how we handle result freshness for queries where a user has chosen a specific date range. This helps ensure that users get the results that are most relevant for the date range that they specify.
• Prediction fix for IME queries: This change improves how Autocomplete handles IME queries (queries which contain non-Latin characters). Autocomplete was previously storing the intermediate keystrokes needed to type each character, which would sometimes result in gibberish predictions for Hebrew, Russian and Arabic.

If you’re a site owner, before you go wild tuning your anchor text or thinking about your web presence for Icelandic users, please remember that this is only a sampling of the hundreds of changes we make to our search algorithms in a given year, and even these changes may not work precisely as you’d imagine. We’ve decided to publish these descriptions in part because these specific changes are less susceptible to gaming.

For those of us working in search every day, we think this stuff is incredibly exciting -- but then again, we’re big search geeks. Let us know what you think and we’ll consider publishing more posts like this in the future.

Posted by Matt Cutts, Distinguished Engineer

The +1 button for Images: See it. Share it.

Recommendations from your friends help making decisions on everything easier, from where to eat, to what movies to watch and where to go for your next trip -- and pictures can really help you decide. In March, we introduced the +1 button to let you make recommendations right from your search results page. Now, we’re extending this ability to Google Images.

Let’s say you’re looking to summit Mount Kilimanjaro and want to inspire a few of your climbing buddies to join you. You search for [mount kilimanjaro summit] and switch to Images mode to find rows and rows of photos testifying that this peak can indeed be conquered. By hovering over one of the images, you can quickly recommend this photo to your friends by clicking the +1 button.


Once they’re on board with the trip and start searching too, you’ll be able to quickly spot images they’ve recommended -- you’ll see annotations on the images they’ve +1’d on your search results page.


As with the other +1’s, you’ll need to create a Google+ profile before you can start +1’ing images. Your image +1’s will appear in the +1 tab of your profile, where you can see all of your recommendations in one place and delete those you no longer feel strongly about. To see +1’s on the image results page you’ll need to be logged into your Google Account.

Posted by Xiaorui Gan, Software Engineer

Powering a new job search engine for military veterans

(Cross-posted on the Official Google Blog and the Public Policy blog)

Earlier today, President Obama spoke about the importance of helping returning military veterans find work. Thousands of businesses have committed to hiring military veterans and families and as part of this nationwide effort, starting today, job seekers can visit the National Resource Directory (NRD) to search more than 500,000 job openings from employers around the country.

We have been working with the U.S. Department of Veterans Affairs to provide a customized job search engine for the NRD, using Google Custom Search technology. This custom search engine uses the power and scale of Google search to constantly crawl the web, looking for JobPosting markup from Schema.org on sites like simplyhired.com to identify veteran-committed job openings. An employer can easily add a job posting to NRD simply by adding that markup to their own web page. As pages are updated or removed from the web, they’re automatically updated and removed from the system, keeping the available job postings on NRD fresh and up to date.

If you’re an employer, you can find more information on how to participate on nationalresourcedirectory.gov. In addition, organizations such as local veterans' groups can help people find jobs by adding a veteran-committed jobs search box to their websites. 

We’re happy to contribute to this important initiative and hope businesses use this opportunity to connect with veterans seeking employment. 

Posted by Christina Chen, Product Manager, Search

Giving you fresher, more recent search results

(Cross-posted on the Official Google Blog)

Search results, like warm cookies right out of the oven or cool refreshing fruit on a hot summer’s day, are best when they’re fresh. Even if you don’t specify it in your search, you probably want search results that are relevant and recent.

If I search for [olympics], I probably want information about next summer’s upcoming Olympics, not the 1900 Summer Olympics (the only time my favorite sport, cricket, was played). Google Search uses a freshness algorithm, designed to give you the most up-to-date results, so even when I just type [olympics] without specifying 2012, I still find what I’m looking for.

Given the incredibly fast pace at which information moves in today’s world, the most recent information can be from the last week, day or even minute, and depending on the search terms, the algorithm needs to be able to figure out if a result from a week ago about a TV show is recent, or if a result from a week ago about breaking news is too old.

We completed our Caffeine web indexing system last year, which allows us to crawl and index the web for fresh content quickly on an enormous scale. Building upon the momentum from Caffeine, today we’re making a significant improvement to our ranking algorithm that impacts roughly 35 percent of searches and better determines when to give you more up-to-date relevant results for these varying degrees of freshness.

• Recent events or hot topics. For recent events or hot topics that begin trending on the web, you want to find the latest information immediately. Now when you search for current events like [occupy oakland protest], or for the latest news about the [nba lockout], you’ll see more high-quality pages that might only be minutes old. 

• Regularly recurring events. Some events take place on a regularly recurring basis, such as annual conferences like [ICALP] or an event like the [presidential election]. Without specifying with your keywords, it’s implied that you expect to see the most recent event, and not one from 50 years ago. There are also things that recur more frequently, so now when you’re searching for the latest [NFL scores], [dancing with the stars] results or [exxon earnings], you’ll see the latest information. 

• Frequent updates. There are also searches for information that changes often, but isn’t really a hot topic or a recurring event. For example, if you’re researching the [best slr cameras], or you’re in the market for a new car and want [subaru impreza reviews], you probably want the most up to date information. 

There are plenty of cases where results that are a few years old might still be useful for you. [fast tomato sauce recipe] certainly saved me after a call from my wife reminded me I had volunteered to make dinner! On the other hand, when I search for the [49ers score], a result that is a week old might be too old.

Different searches have different freshness needs. This algorithmic improvement is designed to better understand how to differentiate between these kinds of searches and the level of freshness you need, and make sure you get the most up to the minute answers.

Update 11/7/11: To clarify, when we say this algorithm impacted 35% of searches, we mean at least one result on the page was affected, as opposed to when we've said noticeably impacted in the past, which means changes that are significant enough that an average user would notice. Using that same scale, this change noticeably impacts 6 - 10% of searches, depending on the language and domain you're searching on.

Posted by Amit Singhal, Google Fellow

Highlighting journalists on Google News

(Cross-posted on the Google News Blog)

Great journalism takes more than facts and figures -- it takes skilled reporters to knit together compelling stories. Knowing who wrote an article can help readers understand the article's context and quality, see more articles by that person, and even interact directly with them. Whole communities can form around prominent contributors, which is why we started showing information about content creators next to their material in Google Search.

Accordingly, Google News is rolling out more information about journalists over the next several weeks, starting with English-language editions. When reporters link their Google profile with their articles, Google News now shows the writer’s name and how many Google+ users have that person in their circles. For the lead article for each story, Google News also shows that reporter’s profile picture and enables readers to add them to their Google+ circles right from the Google News homepage.

If you are a journalist and would like to participate, please follow the instructions in our Help Center. If you are a reader, we hope you enjoy learning more about the faces behind the news.

Posted by Eric Weigle, Software Engineer, Google News team

A new way to get information about real-world places

I love eating out with my friends and trying new places, but one of the most difficult questions you can ask me is "Where do you want to eat?" Today, we’re making a few improvements to Google search that will make it a lot easier and faster to answer this question. For example, I mentioned to a friend that I’ll be visiting Boston, and he suggested that I check out a barbecue place called Redbones BBQ in Davis Square. Since I don't know that restaurant, I do a quick search for [redbones bbq] to see if it’s a place I’d like.


When I do, I see the same familiar search results page but I notice that there's now a new panel to the right of the results -- where previously only a map appeared -- with much more information than before. I see two images with pegman, the Street View mascot, below the map so I click on the first one. This instantly takes me to an immersive 360-degree interior view of the restaurant, as if I were virtually teleported to Redbones.


I pan around and see that it's a cool colorful restaurant with a nice, comfy feel. What's more, when I go back to the search results panel and click on the second image, I’m able to look around the outside of the restaurant and get a sense for the neighborhood via the familiar Street View experience. I’m beginning to really like this place! Further down the panel, I see the price range indicating it won’t be too expensive and an “at a glance” summary that tells me Redbones has great beer and pulled pork sandwiches -- and menu links if I want to see more. Thanks to this helpful information right on the search results page, I’ve quickly been able to make my decision: I’m going to Redbones for a pulled pork sandwich.

Even if I’m not looking for a particular place by name, I can learn about places and quickly decide which ones are right for me. If I want to find a bar near Redbones for a few drinks after dinner, I can just search for [bars davis square] and get a familiar list of results. Only now, scanning the list and comparing places is easier than ever, since the instant preview feature will show the same detailed information about the various bars when I hover over the “>>” symbol to the right of each result. After just a few seconds perusing the additional local information for different places, I know that Joshua Tree has a great beer selection and that the Orleans has live music but is a bit farther away.


This new type of layout may appear on the search results page for a range of real-world places -- restaurants, hotels, local businesses, landmarks, museums and more. Of course, the local information that appears will vary depending on what’s available online. So the next time you plan your visit to the New England Aquarium or Fenway Park, you might be able to check out their opening hours, get directions, and find the nearest transit stops, all from a simple Google search.

In the coming weeks, you’ll start seeing the improved local search experience in more than 40 languages. Give it a try and start discovering new local favorites, near and far!

Posted by Jim Müller, Tech Lead