Beta Labs - Yellowpages.com

AT&T Interactive (ATTi) is recognizing the spirit of the Olympic Winter Games through a new YP.COM Review2Win Sweepstakes.

To enter, simply write a review about your favorite businesses on the new YP.COM Beta site and you and a guest could soon be headed to Vancouver to Cheer on Team USATM at the 2010 Olympic Games! Enter the Review2Win Sweepstakes and write reviews daily for additional chances to win a daily prize, and also try your luck daily at our instant-win game where you’ll have a chance to score some of our hundreds of Visa Gift cards.

$25-$500 Instant-Win Opportunities – Play Daily!
Create a profile on YP.COM Beta and write a review about a favorite business. Then register to play — if you get three matching Olympic Winter Games sports figures, you’ll instantly win a Visa Gift card valued between $25 and $500! You can review as many businesses as you like, but are limited to one entry per day.

Enter the Grand Prize Sweepstakes
Register for the sweepstakes to get an opportunity to win the Grand Prize — a trip for two to the Vancouver 2010 Olympic Games. The random drawing will be held at the end of the promotion period. Log in and write reviews once each day to earn another entry into the Review2Win Sweepstakes.

The contest ends December 31, 2009 at 11:59 p.m. (ET) so don’t wait for your chance to win great prizes and to be part of the Vancouver 2010 Olympic Games.

Spread the Word
A “Tell a Friend” function allows you to share the Review2Win Sweepstakes with your friends and family to give them a chance to win, too. So every time you play, “Tell a Friend.”

See the new YP.COM Beta site for complete official contest rules.

Ratings and Reviews on YELLOWPAGES.COM

Last week we launched the Review2Win Sweepstakes on yp.com and YELLOWPAGES.COM. The sweepstakes should generate increased awareness of our ratings and review feature (games, prizes and trips are gauranteed to get our attention!) so we decided that this was a great opportunity to introduce Nicholas Perry – the ratings and reviews feature manager. We asked him to give us an overview of the feature and highlight why ratings and reviews matter in local search.

“People generally trust online reviews more than critic reviews and editorials. They want to read what other people thought about a business, learn why others rated a business highly or poorly, and get a better idea of whether they want to use that particular business or not. Ratings and reviews help a user differentiate between businesses that otherwise seem very similar.

On YELLOWPAGES.COM and yp.com we allow users to rate businesses on a scale of 1-5 stars and write a review of their experience with a business. We also aggregate reviews from other sites to increase the number of ratings and reviews that we provide to our users.

On the yp.com site, we use the reviews to help provide relevant results. If many users are praising a restaurant’s lasagna in their reviews, then that business may be a better match for a user searching for “lasagna” than other Italian restaurants

We filter reviews* for violations of our specific guidelines, such as profanity, slander, or spam. If a user writes an honest, factual review of a bad experience they had with a business, then that review will remain on our sites. We can’t control everything, but the more real users we have reviewing each business, the less impact fake reviews will have.

We present the reviews in order of review date. Recent experience with a business is most relevant to how the business is performing now, helping a user make a decision on which business to use.

We recognize the importance of including businesses in the ratings and reviews process. We allow businesses to reply directly on the site to each review so business owners can have an open dialogue with their customers for positive or negative reviews. This dialogue between consumers and the business can be very valuable to everyone. And, since we recognize that some businesses are also concerned about the quality of reviews we do allow businesses to opt out of reviews altogether. However, as online reviews are extremely valuable and important to consumers, the vast majority of businesses do allow users to review their business.

There are many prizes available during our Review2Win Sweepstakes, so take this opportunity to provide very valuable insight into local businesses that you love or hate. Your voice will be heard by our large user base and may help your neighbors make good decisions on which local businesses to use.

Watch for new innovations with our Ratings and Reviews feature in the future. We are currently in the process of developing new functionality which we plan to introduce in 2010. Of course, if you have any ideas that you’d like to see on the site, just drop us a note. All ideas are welcome!

Nicholas Perry
Sr Product Manager

* “We are under no obligation to monitor any reviews; however, we reserve the right, in our sole discretion, to monitor, edit and/or remove any reviews for any reason, investigate possible violations of our User Review guidelines, and enforce any rights available to us as a result of any such violation.”

New hotel widget on YELLOWPAGES.COM – just in time for holiday travel!

With over 39% of all travel bookings being made on the web*, consumers have high expectations of online travel resources. We believe we can meet these expectations and deliver a more satisfying travel planning experience with the new hotel rates and availability search widget, which will launch on YELLOWPAGES.COM and YP.com in November and December 2009 respectively.

With the widget consumers can easily check hotel rates and availability, and book a room. If users find their favorite hotel completely booked for their vacation dates, they can also search for other options through the widget.

Traveling during the holidays can be stressful, so relax while you’re planning: slip into your favorite flannel jammies, get a cup of hot cocoa and use the YELLOWPAGES.COM hotel rates and availability search widget to help plan your trip. Afterwards, drop us a note and let us know how it worked (the widget, not the cocoa). This is just our first phase, so stay tuned for more from your Hotels Vertical Team.

Steven Lee
Sr. Product Manager for Hotels Vertical

* PhoCusWright’s U.S. Online Travel Overview Ninth Edition

m.yp.com mobile website demo with Stephen Chappell

In the following video Stephen Chappell (our Director of Mobile Products) gives a short, informative demo of m.yp.com.

Watch the video, try the site – and let us know what you did or didn’t like.

Stephen is here at the office in Glendale so we can always put him in front of the camera if you need more info.

Local Search, Data Accuracy and the Elusive Canonical Phone Number

An Interview with Neil Symes - Director of Data Management

ca•non•i•cal (kə-nŏn’ĭ-kəl)
1. Of, relating to, or required by canon law.
2. Of or appearing in the biblical canon.
3. Conforming to orthodox or well-established rules or patterns, as of procedure.
source:
www.thefreedictionary.com

One of our jobs as a local search engine is to provide accurate and up-to-date information – including business telephone numbers. This sounds like a simple, straight forward proposition, but it is not. In fact, this is extremely difficult due to the complicated relationship between businesses and their telephone numbers. Our Director of Data Management, Neil Symes, spends a great deal of effort pursuing the holy grail of the canonical phone number.

Q: Can you give me a brief description of what we call a “Canonical” phone number?
A: Well, in a nutshell, it’s the actual phone number for that business. When we list a business, we want the phone number to be the ACTUAL phone number the person will need to call to reach the business, not their fax number or back office number.

Q: Why is this data difficult to obtain?
A: There are many, many reasons this is problematic. There are toll free numbers, call tracking numbers, fax numbers, even businesses with multiple numbers in multiple area codes. In addition, a business may have a primary line with hundreds of extensions. These extensions may have direct dial capability as well. So “what is the phone number” for a business, person, or department at a business is extremely complex. The control of phone numbers is split between the traditional Telcos, wireless providers and VOIP so obtaining a standard across the industry for reporting the canonical phone number could be a challenge. With canonical web addresses, the control and benefit are both in control of the website owner; with telephone numbers, the control is within the various telephone service providers and advertising agencies, with the benefits going to the search engines. Even within the Telcos they have difficulty tracking this information. For example, phone numbers are recycled / reassigned quickly when a business closes its doors. In some drastic circumstances this can occur in a 30 day period.

Q: What other challenges are there?
A: Mobile numbers have added an entirely new dimension to the data accuracy question, as they do not point to a location, but to a person. There are also situations where a national business will want different phone numbers displayed, such as a national number on a search results page, but a local number on a MIP (More Information Page).

Q: Is this a problem for all search engines?
A: This is mainly a problem for Local Search web sites. A typical search engine brings back a list of relevant websites, but in local search we bring back a list of relevant businesses and their telephone number. In this context it’s important to note that not all businesses have websites, but most do have a telephone number. Determining the right telephone number for a business is crucial for local search. This isn’t an easy problem to solve, but we apply innovative technology and our deep understanding of businesses and their use of telephone numbers to develop solutions. Are we perfect? No, but the reality is – no one has perfectly solved this problem. We are always looking for better solutions and would love to hear from other “students of the canonical phone number” how to be even better.

User Feedback on Search Relevance

When writing the retrieval & ranking code that’s at the heart of any search engine — code that answers the question what results should be returned for this search, and what order should they be in? — engineers have a plethora of algorithms to choose from. And given the critical importance of getting relevance right, one might reasonably wonder whether, by now, the search industry has settled on a standard automated measurement of relevancy. (That would certainly help us choose the right search algorithms!)  This is not the case.  While some work has been done to analyze server logs to report proxy measures such as click-through rates and page dwell time, automated tools have yet to replace human raters. This is largely because relevancy is highly subjective.

Many local search engine firms have specialized in-house teams whose job it is to judge the relevance of the results produced by their engine in response to real-world queries from users. But this raises some questions: Are the scores from a group of ‘professional judges’ applicable to the general population? Without local knowledge, can the testers determine whether an ambiguously named business is properly categorized or whether the most popular restaurant in a neighborhood is ranking appropriately?  Are the testers aware of colloquial differences in names for businesses or products?  Do the testers understand local preferences around how far people are willing to travel for a particular type of business?

One way to address such concerns is simply by asking for direct feedback from real users. So today, yp.com is launching a tool to gather relevancy feedback on individual search results. Next to any business listing show on a search results page on yp.com, you’ll notice a pair of Yes/No buttons under the words “Listing relevant?” We hope you’ll take the time to click a few of them — especially when we return the wrong stuff, but also on those occasions when we manage to get it right, too!

Introducing YP.COM

Our mission at AT&T Interactive is to help consumers discover local and to help advertisers grow their business. So today we’re excited to launch YP.COM Beta, our next step toward broadening local search and advertising across online, mobile and social.

As a complement to our YELLOWPAGES.COM property, which focuses on presentation and placement of geo-targeted advertisers, YP.COM Beta sharpens the local search experience by providing consumers with increased relevant local results, while providing advertisers with more relevant targeting. To pull it off, we’ve been busy!

Makes sense, right? Increased relevancy leads to better results. Better results lead to happier consumers and advertisers. It’s just that simple

Our teams built YP.COM Beta from the ground up. From new data systems and a new local search stack to a new user interface and design. We also added some great features to help consumers discover and transact with local businesses:

• Expandable local maps in results
• Improved search refinements
• Aggregated content about local businesses
• Videos, coupons, and promotions from local businesses

We have more features coming and can’t wait to unveil them. So stay tuned and watch the site evolve. In the meantime, give YP.COM Beta a spin, shoot us some feedback and tell us what we can do to help you discover local.

-David Yoo

AT&T Interactive

Chief Product Officer

Alternative Local Search Algorithm Designs

Our Search Engineering team has been researching alternative algorithms for delivering highly relevant local search results for web and mobile users, and we’re pretty excited about our progress.

The existing approaches, in a simplified sense, rely upon matching keywords and/or category names based on just the name, keyword and category data associated with the available business listings. While we’ve implemented a fair number of mechanisms to identify near matches, related matches and possible disambiguation choices, there were things that we felt we could do to search our data in a more effective way.

This current round of development has yielded an approach based on a couple of interesting methods: Free Search and Geo Density. Free Search is our TF-IDF (Term Frequency and Inverse Document Frequency) based algorithm that includes all available data in our listing index and applies weights to fields based on our internal search logic. This gives our searches much greater penetration into our data set than the existing search method, providing for an information retrieval method that is more responsive to how people might casually describe what they are searching for.

If Free Search represents our approach to delivering a better “what” component of our searches, Geo Density is our new approach to improving the “where” component. When we build our indexes of our listing data, we are also calculating the relative density of various listing types in the named geographies we can search in. For this beta example, we are calculating these densities against the geography of California, but we will soon expand this to support a much more granular breakdown of density of information (i.e., the search model will be able to understand that the density of coffee shops in Santa Monica, CA is different than the density of coffee shops in Mojave, CA). This density value allows us to dynamically set the radius of our search and apply a weighting (not a hard filter) to results within that radius.

This sort of smart and dynamic distance determination is key to a friendly local search system, as it provides a more intuitive scaling of the results, based on the availability of the thing being searched for and the specificity of the search.

For example, a search for “furniture” in Pasadena, CA will return a focused and relevant list of results in and around Pasadena, while a search for “IKEA” in Pasadena, CA will automatically scale out to include more IKEA stores in the results. The search system can do this quickly and efficiently because it understands the relative density of stores of this type in California and can infer the specificity of the search from the search term used.

As another example, take this search for “coffee shop” in Pasadena, CA and compare it to a similar search for “Starbucks” in Pasadena, CA. Both searches remain well focused on the intended geography, though the area will change slightly due to the specificity of the search term used.

It is important to note that, as can be seen in the examples above, businesses of related categories sometimes occur in the result sets. This is due to the way the system includes and weights the first few categories associated with the initial query. This expansion of the search domain is an experiment to determine whether or not we can provide a more intuitive scope for searches, and will be evolving as we update the system in the near future.

We will be updating the engine in the near future to increase its accuracy and sensitivity, so please leave us feedback and check back soon for an update.

Note: Only California listings are included in our testing at this time.

Improving Mobile Search for Cities

Our first-generation mobile search algorithms often use simple “expanding radius searches” around specific points. Here’s what happens:  If you were to conduct a search for pizza in Chicago, IL, we would search within a 5-mile-wide circle around a central point in Chicago for businesses related to pizza.  If no results are found we would expand the radius by an additional 5 miles, until we find results or reach the maximum search radius of 20 miles. The same goes for ZIP codes, neighborhoods and street addresses.

The problem is that for some cities (not so much for ZIP codes and neighborhoods), this simple approach can sometimes omit businesses that actually fall within the city, but don’t fall within the circle we’ve drawn on the map. This is especially perplexing for users who search for a business that they know is in a certain city, only to discover that they can’t find it using one of our mobile apps.

A simple improvement to this algorithm is to combine the basic idea of a radius search with a search for businesses with addresses that match the city name. We call this search algorithm “Box+ search”.  We expect this new dual approach to provide more accurate search results than the “expanding radius search”.  To validate this we have set up a page where you can see the results of the two approaches side by side.  You can see the difference in search results by looking at the following links, or try ‘Box+ Search” search with your own terms and location:

• Thai Food in Pasadena, CA
• Apple Stores in New York, NY
• Day Spa in Malibu, CA*
• Auto Parts in Sierra Madre, CA
• Massage in Sammamish, WA

Note: For uninteresting technical reasons, the search actually finds matching businesses using a box, rather than a circle, along with the name of the city, neighborhood or zip code.

The blue pins are businesses that were only returned in the “expanding radius search”.  The red pins are businesses returned with “Box+ search”.  The green pins represent listings that were returned by both search algorithms.

The red box shows the default area that was searched for Box+, while the blue circle shows the expanded radius search area.

* This result is interesting because it shows how a Box+ Search returns accurate results i.e. business that are “Day Spas” in “Malibu, CA”, where as the radius search returns businesses in similar categories but closer to the center of the specified search area.

Single Search Box

As we mentioned in Where You At?, it is an unfortunate fact of life that not all searches arrive at the yellowpages.com servers neatly divided into two parts: the what part (“pizza” or “plumbers”) and the where part (“albuquerque” or “anchorage”).

These untidy queries arrive as an undifferentiated jumble of text—e.g. “pizza los angeles ca”—and while it’s obvious to a human what that means, it’s not so perfectly transparent to a computer. So it’s up to our search engineers to write programs that figure out which parts are the what of the user’s query, and which parts are the where.

We have recently done a bit of work on this front, and would like to share an initial implementation; it’s not perfect, but we hope it handles many common cases. (It works by comparing the sub sequences of words in the query string against known addresses in our database; we use the matches in order to distinguish the geographic terms in the query from the rest. Call it separating the where from the chaff …)

If you’d like to try it out, you can use the field below:

As usual, please use the comments to let us know if you find cases where it ought to work, but doesn’t.