Formed in 2006, Microsoft® adCenter
Labs combines the talents of more than 120 top researchers and engineers,
all dedicated to developing cutting-edge new ways for consumers
and advertisers to connect. These researchers work at Microsoft
Corp. facilities in Redmond, Wash., and Beijing across a broad range
of technology areas, including data mining, information retrieval,
statistical analysis, artificial intelligence, auction theory, visual
computing and digital media.
AdCenter Labs researchers are working on
more than 60 online marketing tools and technologies, which are
at various stages of development, testing and integration into Microsoft
advertising products and services. These tools and technologies
include breakthrough innovations in the following areas:
Video-Display Ads
Online video provides advertisers with a
powerful medium to connect with consumers, particularly when the
video includes interactive and other advanced technology.
• Social video sharing. This tool allows
users to highlight and comment on a specific area of a video frame.
This enables consumers to have meaningful social interactions around
specific things in the video, such as being able to discuss how
a goalie just blocked a soccer goal, or ask questions about a step
in a cooking instruction.
• How it works. It uses Javascript
to implement the highlighting and commenting steps. It also allows
for authentication using Windows Live™ ID, so users can choose
to view comments from their friends only.
• Benefits, advantage. Other online video sites do not allow
labeling on a specific area of a frame, but users of adCenter Labs’
tool can. This allows a new level of interaction between users.
• Current status. A demonstration is available at http://adlab.microsoft.com/vad.
• Video hyperlinks. The first stand-alone product developed
by adCenter Labs, this tool lets marketers embed video hyperlinks
within online video ads. The hyperlinks, which can be associated
with specific products in the video, take consumers to Web sites
where they can obtain additional information about products or purchase
them.
• How it works. A computer-vision algorithm
tracks selected objects as they appear in key frames of the video.
An array of information, including text, graphics and links to Web
sites, can be assigned to each object.
• Benefits, advantages. An interactive video experience made
easy by adCenter Labs’ authoring tool can dramatically reduce
the time necessary to generate the digital tracking and information
file through use of computer-vision algorithms.
• Current status. A pilot ad for a national retailer is slated
to appear in spring 2007. A demonstration is available at http://adlab.microsoft.com/vhl.
• Large display feedback. Vision-based technology creates
interactive public displays that consumers can remotely control.
The technology also can measure the size of the audience and estimate
demographics.
• How it works. A computer-vision algorithm
tracks the hand movements of people from a large video screen. By
moving their hands in the air, people can control what happens on
the screen. Face-detection and machine-learning algorithms count
the number of people in the audience and predict the gender.
• Benefits, advantages. Advertisers can engage consumers in
public locations, as well as get a sense of the number and types
of people who have seen their ads in public locations where such
measurements have previously been difficult or impossible.
• Current status. A demonstration is not available at this
time.
Keyword Optimization
Keyword technology is vital to paid-search
and content ads, allowing advertisers to reach consumers who are
interested in their products and services based on the terms they
enter into online search tools.
• Commercial intent detection. Advanced
keyword analysis helps differentiate consumers who are looking to
make an online purchase from those who are only searching for information
about products.
• How it works. An algorithm analyzes
the terms that consumers type into the search engine and compares
queries typically used by online shoppers with those used by people
who are only looking for information but not shopping. The tool
then generates a score that estimates how likely consumers are to
make purchases online. For example, the term “LCD monitor”
carries commercial intention of 0.74 out of a possible 1.0. The
same type of analysis is done on Web sites to help ensure the Web
sites presented to users match their intention.
• Benefits, advantages. It is designed to improve the relevance
of search results and help advertisers optimize their messages and
campaign strategies.
• Current status. A demonstration is available at http://adlab.microsoft.com/OCI/OCI.aspx.
• Query entity detection. adCenter technology is designed
to detect the best way to break down a multiword query. For instance,
if a user searches for “real estate agent,” adCenter
Labs’ tool can tell that the “real estate” is
one phrase and “agent” is another. This technology helps
to determine the meaning of a query term so the relevance of search
results can be improved.
• How it works. The algorithm analyzes
the relationship between words in a search term and decides the
best way to break it down.
• Benefits, advantages. Many search engines treat query terms
as a collection of single terms, often missing the information contained
in a multiword phrase. adCenter Labs’ technology can detect
that “real estate” is one entity and treat it as such
when searching for the most relevant results.
• Current status. A demonstration is available at http://adlab.microsoft.com/adev_demo.
• Keyword service platform. This server product provides a
set of services that can be built into Web applications to analyze
keywords.
• How it works. AdCenter Labs provides
a set of application programming interfaces for Microsoft and, eventually,
for third-party search engine marketers designed to more effectively
suggest, categorize, monetize and forecast traffic and extract keyword
terms.
• Benefits, advantages. The individual services can be combined
in different ways to suit the needs of different developers and
advertisers.
• Current status. A demonstration is available at http://adlab.microsoft.com/KTS/KCAT.aspx.
• Contextual advertising. AdCenter Labs’ technology
classifies Web sites and keywords into hierarchical taxonomy. This
helps increase the relevance of contextual ads by making sure the
categories of publisher sites are relevant to advertisers’
products and services.
• How it works. An algorithm analyzes
the content of a Web page and presents ads with related categories.
• Benefits, advantages. adCenter Labs’ categorization
technology enables a new paradigm in online advertising that is
based on “concepts” instead of keywords. Many advertisers
need to manage campaigns with hundreds of thousands of keywords
to maximize the reach to their consumer audience. Concept-based
advertising is designed to significantly simplify the work of advertisers
and their agencies.
• Current status. A demonstration is not available at this
time.
Consumer Orientation
The placement of online advertising must
factor in the tendencies and tastes of different consumers audiences.
• Sensitive content detection. This
technology helps determine whether a Web page contains content that
the general public might consider objectionable, such as pornography,
crime or terrorism.
• How it works. An algorithm analyzes
the content on a page to determine whether the page contains objectionable
content.
• Benefits, advantages. Advertisers can avoid associations
with objectionable content and delivering commercial messages in
inappropriate contexts. This can also improve users’ search
experiences to help ensure they are sufficiently warned regarding
the content of a Web page they are about to visit.

learn more about these free offers from microsoft:
windows live onecare | windows live toolbar | microsoft adCenter
|