In their motion for disqualification of Deborah Platt Majoras, the Center for Digital Democracy and the Electronic Privacy Information Center cited a newly discovered fact about the case. Majoras’ husband, John Majoras, works for a law firm retained by DoubleClick.

John Majoras is not listed as one of the five attorneys DC law firm Jones Day has advising DoubleClick on antitrust issues. However, searching the Jones Day site, and Google, shows references to DoubleClick no longer appear on the firm’s website.

A query through the Jones Day site search displayed this item as one of four results:

3. Jones Day - Services - Antitrust Mergers/Joint Ventures - Experience
…DTE Energy Company. DoubleClick Inc. Acquisition by DoubleClick Antitrust counsel in the $3.1 billion acquisition…in the $3.1 billion acquisition of DoubleClick by Google Inc. H&R… http://www.jonesday.com/services/services_experience.aspx?
AdvancedServices=7deaa765-88f7-4368-8db7-443695819af3
12/7/2007 1:10:31 AM

Click the link, and DoubleClick does not appear in the resulting list. We expect this could be a simple administrative oversight, and look forward to hearing it has been corrected.

But the Jones Day webmaster will have more work to do. For one thing, this link should lead directly to a summary of Jones Day’s experience with DoubleClick. It’s blank, but you can see it in the CDD/EPIC motion.

We checked each of the five attorneys’ profiles on the site. EPIC and CDD listed Joe Sims, Thomas Jestaedt, Alexandre G. Verheyden, Michael S. McFalls, and Chris Ahern as the legal eagles providing counsel.

Each of those attorneys once had this listed in their Professional Biography – Experience:

DoubleClick Inc.
Acquisition by DoubleClick
Antitrust counsel in the $3.1 billion acquisition of DoubleClick by Google Inc.

In each case, that reference has been scrubbed from the Jones Day site. One would imagine an attorney would want to tout being part of a multi-billion-dollar deal involving one of the biggest name brand companies on the planet.

Not at Jones Day, though.

(UPDATE: A DoubleClick representative responded to this story via email: "Simpson Thacher has been DoubleClick’s outside counsel since July of 2005 and was retained to represent it in all aspects of its proposed acquisition by Google, including with respect to United States antitrust matters.  From the outset, Simpson Thacher has represented DoubleClick before the Federal Trade Commission and continues in that capacity. Jones Day has been engaged primarily with respect to European and other non-U.S. jurisdictions.  Jones Day was not engaged to represent, and has not represented DoubleClick before the Federal Trade Commission or appeared before the Commission on DoubleClick’s behalf.")

follow me on Twitter

 An example: when you enter the query http://googlified.net, Google will return:

No results found for http://googlified.net.

However, there are results from websites with similar addresses:

Googlified
…
googlified.com/
…

At other times, mistyping a URL will return a normal spellchecking result. E.g. entering http://blogoscope.com into Google at this time returns “Did you mean: http://blogoscoped.com”, showing the snippet of blogoscoped.com below.

Ionut Alex. Chitu, who noticed this new behavior, comments on the reason why Google might have introduced this special result type:

Many people don’t know the difference between Google’s search box and the browser’s address bar. This confusion is accentuated by the fact that the address bar is also a search box in some cases and Google, like most search engines, steal the focus from the address bar.

Comments

Tag:

“Recent research conducted by Harris Interactive* indicates that consumers are suffering from "Web Search fatigue." The study revealed that while the vast majority of the population uses a search engine to find information on the Internet, only 15 percent of people find what they’re looking for with their first search. Typically, people need to conduct three to four searches to find the right result.

While search engines can and will continue to improve, we think making queries smarter is also part of the solution. While there are a handful of seasoned users and professional searchers out there who employ Boolean operators, site restrict and other advanced techniques, no one (ourselves included) wants to have to think hard about which search terms to use. Our vision was to build a system that helps users confidently construct more sophisticated and targeted queries without slowing them down.

In this first release, we’ve broken the problem into two parts. Part One provides assistance as you type. This has massively reduced the number of misspelled queries being submitted and saves typing time for slower typists. It can also link you deeper into the results space, saving you a step on those quick in-and-out sessions.

Part Two helps you move your searches forward, backward or sideways so you get the results you actually want more quickly. So if you started with a query for ‘Thom Yorke,’ you might get a forward (refining) suggestion for ‘Eraser,’ his most recent album, or a backward (expanding) suggestion for ‘Radiohead,’ the band he’s a

member of. You could also get a sideways (or related) suggestion such as ‘solo albums‘ which, when clicking on the term in the left-hand column of Search Assist, will lead you to other lead singers with solo albums like Gwen Stefani or Eddie Vedder.”

More info on Yahoo! Search blog.

Comments

Tag:

On the positive side, broad matching means that ads will run for variations of your keywords without the need to specify what all these variations might be. So in the example given in the adwords help file, the keyword web hosting might trigger ads for the search phrases web host or web hosting company.

However broad matching incorporates expanded matching so not only will ads show for variations in word order, and queries with additional words to the keywords, but ads will potentially show for synonyms and other related variations.

This might not be what you want and there may be times when the expanded matching can be much broader than you might expect. In a campaign I recently set up, a broad match on a generic product name triggered ads for search queries for a manufacturer name, very specific model numbers (not relevant to my client) and other products sold by the manufacturer.

There are two main ways to deal with it.

1. Build up a list of negative keywords and add them to the campaign

• use the keyword tools (particularly the adwords keyword tool as this will indicate potential expanded matches); extract the words that are not relevant and including them as negatives in the campaign
• run searches and seeing which words and phrases come up in the search listings and ads to uncover potential negatives
• analyse your website logs to extract non-relevant search queries or words that can be input as negatives
• run the new search query performance report to see which search queries triggered ads and what negatives are necessary in the campaign

2. Opt out of expanded match by using phrase and exact matching

By doing this you know when the ads will be triggered and you eliminate the instances of unwanted expanded matching.

Comments

Tag:

Powerset Seems Powerful

The Powerset Blog gives a good example with a recent “Query of the Week.”  Several spot-on results are shown for the phrase “politicians who died in office,” and Scott Prevost, the company’s Director of Product, writes, “For those who think natural language search is just about asking questions, note that this query isn’t a question or even a complete sentence.”

(Interestingly, a Google search for that same phrase yields a Yahoo Answers page as the number one result, and number two goes straight to the Powerset Blog.)

Prevost later continues, “Note that Powerset also knows that ‘governors’ are types of politicians.  This is an example of how Powerset uses ontologies to identify specific examples that match more abstract concepts.  In the second result, Powerset actually knows that Daniel MacDonald is a member of the Canadian Parliament, and that MPs are types of politicians.”

Yet listening to a company promote itself doesn’t always build trust.  For more perspective, we can turn to Read/WriteWeb’s Phil Butler, who performed a full write-up on pretty much everything that is known so far.  “The big news today is that ‘it is on it’s way,’” he states.

Of course, Powerset has yet to publicly mention a launch date, but would-be users can at least sign up for Powerlabs.  Like other companies’ labs, this should let people get a glimpse of the technology, try it out, and give feedback, as well.

• Jason Burby releases: Actionable Web Analytics: Using Data to Make Smart Business Decisions: Buy it at Amazon
• Dennis R. Mortensen of IndexTools discusses how to Use Web Analytics to determine the width of your Internal Search Query box
• Avinash of Google Analytics’ provides his insightful advice: Excellent Analytics Tip #11: Measure Effectiveness Of Your Web Pages
• Judah at Web Analytics Demystified explains why Google Analytics is still not enterprise level ready.
• Juan Damia discusses the Comparison of churn rates (the rate of attrition, over a period of time, that subscriber-based customers unsubscribe)
• Read about Analytics as Data-Mining is going Mainstream – New York Times posted by Marshall Sponder
• Anil Batra: Web Analyst Interview: Daniel Shields

Comments

Tag:

The first thing we do is to check if the query string exists in the first place before we start using it. It could look like this:

if (Request.QueryString["id"] != null)
{
  // Do something with the querystring
}

The only problem with the above check to see if the query string is null, is that we don’t take into consideration if the query string is filled or not.

That could lead to unhandled exceptions in the code. Instead we should check for query strings like this:

if (!String.IsNullOrEmpty(Request.QueryString["id"]))
{
  // Do something with the querystring
}

Then there is the data type of the query string.

Our code might need an integer of 5 digits to get the right information from the database, so if we pass it a string we could end up with a data type mismatch exception.

So we do the check again more thoroughly:

if (!String.IsNullOrEmpty(Request.QueryString["id"]) && Request.QueryString["id"].Length == 5)
{
  // Do something with the querystring
}

Now we know that we get a query string suitable for further processing.

You can then do more precise data type checks using the TryParse method of most value types or by some other logic.

Comments

Tag:

Add to Del.icio.us | Digg | Reddit | Furl

Bookmark WebProNews:

Mads Kristensen currently works as a Senior Developer at Traceworks located
in Copenhagen, Denmark. Mads graduated from Copenhagen Technical Academy with a multimedia degree in
2003, but has been a professional developer since 2000. His main focus is on ASP.NET but is responsible for Winforms, Windows- and
web services in his daily work as well. A true .NET developer with great passion for the simple solution.

http://www.madskristensen.dk/

Consider the URL example.com/?user=123&account=456 and then imaging what a hacker could do with it. Security or not, sometimes you just don’t want the visitors to see all the query strings for whatever reason.

In those cases it would be nice if we could encrypt the entire query string so it wouldn’t carry any readable information. The problem with one big encrypted query string is that we would break all the code that referenced the query. Code like Request.QueryString["user"] would no longer work, but as usual ASP.NET has the answer to that problem.

What we need is an HttpModule that can turn the encrypted query string into a normal readable one, so that we can still use our old logic like Request.QueryString["user"]. In other words, we want the user to see this

?enc=VXzal017xHwKKPolDWQJoLACDqQ0fE//wGkgvRTdG/GgXIBDd1

while your code sees this

?user=123&account=456.

The HttpModule

The module we need for this task must be able to do a few simple things. It must be able to encrypt the regular query string so that all your current links will automatically be encrypted. It must also be able to decrypt it again so that you can write the code as you normally would. It must also provide a method for encrypting a regular query string if you don’t want to use automatic encryption.

The most important feature of the module is to make it totally plug ‘n play. You should be able to apply the module to any existing website and automatically have query string encryption and decryption without changing any of your code.

Implementation

Download the QueryStringModule.cs below and put it in the App_Code folder of your website. Then add the following lines to the web.config’s section:

<httpModules>

</httpModules>

Because automatic encryption is not always desirable the module has a comment that tells you how to turn it off. The module is well commented and should be easy to modify for any ASP.NET developer.

Example

You can encrypt query strings by using the Encrypt() method of the module from any web page or user control.

string query = QueryStringModule.Encrypt("user=123&account=456");

Then just add the encrypted query string to the links that need encryption. You don’t need to use the method if you use automatic encryption.

Download

QueryStringModule.zip (1,55 KB)

Comments

Tag:

Reddit | Furl

Bookmark WebProNews:

Mads Kristensen currently works as a Senior Developer at Traceworks located
in Copenhagen, Denmark. Mads graduated from Copenhagen Technical Academy with a multimedia degree in
2003, but has been a professional developer since 2000. His main focus is on ASP.NET but is responsible for Winforms, Windows- and
web services in his daily work as well. A true .NET developer with great passion for the simple solution.

http://www.madskristensen.dk/

In all, 60 new technologies were unveiled at Demo Fest. But the really buzzworthy ones, if the hype swirling out of Redmond like a great big smug cloud is to be believed, have to do with search (and video, to be addressed later). Here’s a breakdown of the adCenter tools Microsoft wants you to know about.

Keyword Services Platform

A set of Web service APIs for building intelligent applications for online advertising, including keyword recommendation, forecasting, categorization, and monetization. The KSP is intended to help third-party search engine marketers build their campaigns.

Commercial Intent Detection

Keyword analysis that differentiates between buying consumers and researching consumers. The CID algorithm compares search queries with terms most often used by online shoppers versus terms used by those just looking. The tool generates a score that estimates likelihood to buy.

For example, Microsoft says the term “LCD monitor” carries commercial intention of 0.74 out of a possible 1.0

Sensitive Content Detection

A technology that determines whether a Webpage contains potentially objectionable content like pornography, crime, or terrorism. Automatic identification of such material can allow advertisers to avoid their messages being associated with it. Microsoft says it can also help the user experience by warning searchers about content in the SERPs.

Query Entity Detection

A smart technology said to understand multiword queries and word pairings. “Real estate agent,” for example, is different than “real estate.”

Contextual Advertising

What Microsoft calls “a new paradigm,” isn’t that new and is what all the major competitors are shooting for (perhaps like many of these unveiled offerings). AdCenter Lab’s categorization technology is expected to simplify the work of online advertisers by grouping queries into concepts rather than keywords.

Tag:

Add to Del.icio.us | Digg | Reddit | Furl

Bookmark WebProNews:

One of the fun things that I found working in ColdFusion is the many ways you can manipulate xml data. I’m certain this type of process isn’t exclusive to ColdFusion, but I wanted to put it out there anyway. Many developers have a little apprehension about working with new technologies at first, or trying something that hasn’t been already tried and true and of course documented heavily. And frankly it’s hard to find the extra time with a heavy work load and deadlines to meet to go off experimenting But I managed to put aside some time to look at parsing XML into an object that developers are all familiar with, the Query object. As a developer, writing SQL is common place, so I figured maybe looking at a hybrid approach to XML and SQL in ColdFusion, might prove to offer some benefit.

So I’ll show an example of reading in a Phonebook xml document. It will look like this: You can copy this and save it off as phonebook.xml if you like.

<?xml version="1.0"?>
<phonebook>
<contact category="friend">
<firstName>John</firstName>
<lastName>Smith</lastName>
<phone>412-555-1212</phone>
<email>johnsmith@email.com</email>
</contact>
<contact category="friend">
<firstName>Jane</firstName>
<lastName>Smith</lastName>
<phone>412-555-1212</phone>
<email>janesmith@email.com</email>
</contact>
<contact category="enemy">
<firstName>Bob</firstName>
<lastName>Jones</lastName>
<phone>412-555-1213</phone>
<email>bob-jones@mailserver.com</email>
</contact>
<contact category="co-worker">
<firstName>Bill</firstName>
<lastName>Johnson</lastName>
<phone>412-555-1214</phone>
<email>bill.johnson@someserver.com</email>
</contact>
<contact category="friend">
<firstName>Jack</firstName>
<lastName>Robinson</lastName>
<phone>412-555-1215</phone>
<email>www.jackrobinson.com</email>
</contact>
</phonebook>

Reading in the xml is short and sweet simply by using this little block of code.

<cfhttp
url="http://www.mywebsite.com/myxmldocs/phonebook.xml"
method="GET" resolveurl="No" ></cfhttp>
<cfset mydoc = XmlParse(CFHTTP.FileContent)>

Now you’ve got the xml object in a defined variable name called “mydoc”.

Next you set a variable to contain the child element nodes of the xml document and do the same to define the size of the document, which you will see in a bit as to what it’s used for.

<cfset pb = http://mydoc.phonebook.XmlChildren>
<cfset size = ArrayLen(pb)>

Now create a query object with the phonebook data. As you can see we already know what the node properties and names are to create the myquery object containing the xml “column names”. I say column names to relate the data to a structure like a database.

<cfset myquery = QueryNew("cat, firstName, lastName, phone, email")>

Now that we have the “Columns” of the query object, we use the Size object to give us the parameters to define the Rows like this.

<cfset temp = QueryAddRow(myquery, #size#)>

Now you’ve established a Query Object ‘temp’.

Now comes the part where you fill the Query Object with the data from the XML document. Loop through every “row” in the object adding its value from the xml object like an array.

<cfloop index="i" from = "1" to = #size#>
<cfset temp = QuerySetCell(myquery, "cat",
#mydoc.phonebook.contact[i].XMLAttributes['category']#, #i#)>
<cfset temp = QuerySetCell(myquery, "firstName",
#mydoc.phonebook.contact[i].firstName.XmlText#, #i#)>
<cfset temp = QuerySetCell(myquery, "lastName",
#mydoc.phonebook.contact[i].lastName.XmlText#, #i#)>
<cfset temp = QuerySetCell(myquery, "phone",
#mydoc.phonebook.contact[i].phone.XmlText#, #i#)>
<cfset temp = QuerySetCell(myquery, "email",
#mydoc.phonebook.contact[i].email.XmlText#, #i#)>
</cfloop>

Essentially what you’ve just done is take the xml and turn it into the same type of data structure you would have when you query a database.

The rest should look familiar to many ColdFusion developers who use ColdFusion Components which is a great practice.

Call the query function using the CFINVOKE. This particular function is if you were to sort the phonebook individuals by last name.

<cfinvoke component="pbook_meths" method="sortLName"
returnVariable="Result">
<cfinvokeargument name="q_obj" value="#myquery#">
</cfinvoke>

The output of the result from the function would be no different than that of a normal SQL query.

<table border=1 width=500 align=center>
<th>category</th><th>fist name</th><th>last
name</th><th>phone</th><th>email</th>
<cfoutput query="Result">
<tr>
<td>#cat#</td><td>#firstName#</td>
<td>#lastName#</td> <td>#phone#</td>
<td>#email#</td>
</tr>
</cfoutput>
</table>

The function being called would be stored in its own file as components are in ColdFusion and would look like this.

<cffunction name="sortLName" access="remote"
returnType="query">

</cffunction>

Your component can contain numerous functions manipulating the xml query object just the same as you would a normal query from a database. You don’t have the power that a relational database offers of course, but if you’re comfortable writing SQL, you may find yourself hitting the ground running with working with xml in ColdFusion.

Thanks and Happy Coding.

Tag:

Add to Del.icio.us | Digg | Reddit | Furl

Bookmark WebProNews:

Ben Cortese is a developer and business
analyst for the financial industry and develops affiliate
powered websites. http://www.inetsite.net Copyright 2007.