Data Mining: Text Mining, Visualization and Social Media

Comparing Google and Microsoft's Mapping Coverage

I wrote recently about Google's applaudable improvements to their mapping system, remarking that they are still trailing other mapping services (such as Microsoft's). My observations about the data coverage of Google's mapping is based on many interactions with the system, but it is not remotely scientific.

For example, I wrote about the complete absence of Cozumel - a post which received a comment from Andrew Hitchcock about the missing island of Hong Kong.

Someone going by the name of Jon posted this useful comment on my latest post on the topic:

Trailing it's competitors? What are you smoking? Microsoft doesn't have more coverage, and never did.

It is an interesting comment - particularly to me, after having thought about and read about the power of brands to help users fantasize about the quality of service. Jon is, of course, not very well informed.

However, I thought it might be a good time to consider a more methodological approach. Taking a list of countries ordered by population, and starting at the bottom, I reviewed the data found on Google and Live (Microsoft). I scored each location for the presence of the shape of the country, the presence of populated places (PPL) - that is to say cities, towns and so on, the presence of roads, and which system had the better imagery - meaning the highest resolution, not the most up to date.

In the table below, the presence of features is marked for Google and Live. For example, Y/N means that Google has the feature and that Live doesn't.

Country	Population	Shape	PPL	Roads	Best Imagery
Pitcairn	50	Y/Y	N/N	N/N	Live
Vatican City	800	Y/Y	Y/Y	Y/Y	Live
Tokelau	1400	N/Y			Google
Niue	1600	Y/Y	N/Y		Google
Falkland Islands	3000	Y/Y	N/Y	N/N
Montserrat	5900	Y/Y	N/Y	N/Y	Google
Saint-Pierre and Miquelon	6125	Y/Y	N/Y	N/Y	Google
Saint Helena	6600	N/Y	N/Y	N/Y	Live
Saint-Barthélemy	6852	Y/Y	N/Y	N/Y	?
Nauru	10000	Y/Y	N/Y	N/Y	Live
Tuvalu	11000
Anguilla	13000	Y/Y	N/Y	N/Y	Google
Cook Islands	13000	Y/Y	N/N	N/N	Google
Wallis and Futuna	15000	Y/Y	N/Y	N/Y	Google
Palau	20000	Y/Y	N/Y		Live
British Virgin Islands	23000	Y/Y	N/Y	N/Y	Google
Turks and Caicos Islands	26000	Y/Y	N/Y	N/N	Google
Gibraltar	29000	Y/Y	Y/Y	Y/Y	Google
San Marino	31000	Y/Y	Y/Y	Y/Y	Google
Monaco	33000	Y/Y	Y/Y	Y/Y	Google

It is interesting to note the trend indicated by this summary analysis: Microsoft is wining on country and road data, while Google is ahead with imagery. Of course, there are many other ways to sample the data (e.g. via areas with high tourist traffic), and it would be nice to actually build a comprehensive version of the table above - or even display graphically coverage for different system.

I'll finish with this beautiful image from Google of Tokelau, a tiny nation in the south Pacific (and by nation, I mean non-self-governing colonial territory of New Zealand).

Zombies, Cowboys and Dragons

According to BoxOfficeMojo, the top 3 movies right now are Resident Evil, 3:10 to Yuma and Dragon Wars.

It is interesting to note the timing of the 3rd Resident Evil movie.

On July 11th, there was quite a bit of blog reaction to the new trailer for the 5th installation of the game series, the new Umbrella Chronicles shooter game as well as a new Wii control system - all presented at the E3 conference.

NSF: Science & Engineering Visualization Winners

The NSF has posted the winners to this years competition. My favourite is probably:

Möbius Transformations Revealed
Credit: Douglas N. Arnold & Jonathan Rogness, University of Minnesota, Twin Cities

Any real numbers [sic] can be plotted on a line that runs from negative to positive infinity, but throw in an imaginary component and the line becomes a plane, where complex numbers are plotted on both the real and the imaginary axes. Möbius transformations are mathematical functions that send each point on such a plane to a corresponding point somewhere else on the plane, either by rotation, translation, inversion, or dilation. It may sound confusing, but after watching this simple and elegant explanation of Möbius transformations created by Douglas N. Arnold and Jonathan Rogness of the University of Minnesota, Minneapolis, everything becomes clear. Set to classical music, the video demonstrates the transformations in two dimensions but then backs away and adds a third—placing a sphere above the plane and shining light through it. As the sphere moves and rotates above the plane, suddenly all the transformations become linked, in a way that conveys visually in minutes what would otherwise take "pages of algebraic manipulations" to explain, says Rogness.

Anticipating Halo 3

Halo 3 is being touted as not just the biggest launch for a video game, but also the biggest launch of anything. Have a look at a comparison of the Halo brand with mentions of Resident Evil (the top release from last week) and The Bourne Ultimatum (one of the largest films of the summer).

Next up: beating the iPhone?

 

Search Quality is Brand Quality

Greg Linden points to 'The Effect of Brand Awareness on the Evaluation of Search Engine Results', a paper by Bernard J. Jansen et al:

In this paper we investigate the effect of search engine brand (i.e., identifying name that distinguishes a product from its competitors) on evaluation of system performance. Our research is motivated by the large amount of search traffic directed to less than a handful of Web search engines, even though many are of equal technical quality with similar interfaces. We conducted a laboratory experiment with 32 participants measuring the effect of four search engine brands while controlling for the quality of search engine results. Based on average relevance ratings, there was a 25% difference between the most highly rated search engine and the lowest, even though search engine results were identical in both content and presentation. We discuss implications for search engine marketing and the design of empirical studies measuring search engine quality.

Greg summarizes some of the implications of branding with respect to competition in the search space. However, one might also consider branding in the context of a single provider. Google provides search services in at least the following spaces:

• Web search
• Blog search
• Enterprise search (Google Appliance)
• Video Search (Google Video and on YouTube)
• Gmail
• Local Search

While Google has enjoyed a strong brand initially due to some fundamental changes that they brought to search technology (most noticeably PageRank), the technologies that underlie these services are not all the same and, therefore, provide Google with more brand derived value. For example, enterprise intrawebs do not have the same qualities of structure that PageRank requires; searching on YouTube has fundamental quality flaws (ever had results for a search only to find the the video doesn't actually exist!).

Consider the challenge ahead of companies like Powerset and Hakia - which are attempting to bring another fundamental shift to search. Much of the criticism has been leveled at assumed issues with the technology. However, this is not the only battle ground. Establishing brand where there is none is a huge barrier to entry.

Hakia recently launched a new feature which allows for the comparison of their search results with those from a competitor (Google, Yahoo or Microsoft). Hakia has always been about transparency (compare this with others in the NLP search space), and this feature is another step on the path to engaging users to ease them into brand recognition. Entrants have to be far more creative at this level. Personally, I very much want a PowerMouse tshirt from Powerset.

Conversational Marketing

After my initial post, some contributions have been made to the Wikipedia definition of Conversational Marketing. However, I'd like to see some input from the likes of Pete Blackshaw, Steve Rubel, John Battelle and John Gartner.

What is Conversational Marketing?

There have been a few articles posted recently about the topic of Conversational Marketing. I suspect that there is confusion (I'm confused, at least) due to the head-on collision between the commonly accepted meaning of the word 'conversation':

2 a (1): oral exchange of sentiments, observations, opinions, or ideas (2): an instance of such exchange : talk <a quiet conversation> b: an informal discussion of an issue by representatives of governments, institutions, or groups c: an exchange similar to conversation

and the inherently asymmetrical nature of 'marketing':

1 a: the act or process of selling or purchasing in a market b: the process or technique of promoting, selling, and distributing a product or service

2: an aggregate of functions involved in moving goods from producer to consumer

Perhaps its time for a wikipedia entry on the topic.

The Simpsons

Redefining Conversation

While everyone is enjoying the hoopla around a new web based implementation of a very old approach to presentation, Peter Hirshberg over at Technorati has written a long post on the topic of 'conversational marketing.' This is going to take a while to get into. However, one initial perspective - in fact, something that has been bugging me for a while - is this notion that weblogs are conversations. Let's review the facts: weblogs are not conversations.

A weblog is a means to broadcast (distribute identical content to many consumers). Conversations are symmetric interactions (first you, then me, then you, ...). Weblogs are a new form of interaction - a hybrid of dialogue and monologue presented on a platform which by intention and accident has evolved into a hybrid of channels (do you reply on a comment, in an email, on your blog, ...?) Unfortunately, this new form of interaction has not been adequately modeled.

Bloggers must be very sensitive to the effects of lexical imperialism. Terms like 'consumer generated media' establish an us and them perspective of the intentions behind the content that we generate. Similarly, Doc Searls' statements are worrying:

• The purpose of conversation is to create and improve understanding, not for one party to "deliver
  messages" to the other. That would be rude.
• There is no "audience" in a conversation. If we must label others in conversation, let's call them
  partners.
• People in productive conversation don't repeat what they're saying over and over. They learn from each other and move topics forward.
• Conversations are about talking, not announcing. They're about listening, not surveying. They're about paying attention, not getting attention. They're about talking, not announcing [sic]. "Driving" is for cars and cattle, not conversation.

'There is no audience in a conversation.' I agree with this, however there is an audience for a blog. Labeling people in a conversation a 'partner' suggests equality. But as this applies to marketing it is the wrong suggestion. A partner doesn't try to get you to buy stuff you don't need/want. The implication that the blogosphere is a conversation; that we are all partners; therefore people marketing to us in this 'conversation' are our partners is creepy. Another point to note is that there is a backchannel in the blogosphere. Many of us get emails requesting this that or the other get some exposure. Conversations are transparent to all participants.

Those interested in style may be entertained by the 'conversations don't repeat...' followed by the repeated structure of 'Conversations are about A not B', not to mention the actual repeating of 'they are about talking, not announcing.'

Technorati has tried on and off for a while to figure out what their real business is. This post hints at a model based on 'conversational marketing.'

Google Rushes Map Data

Though Google, via an acquisition and their trademark KIS design aesthetic brought mapping to a wider audience, they have, for a considerable time now, been trailing all of the major players (most noticeably Microsoft whose online mapping service has a lot more coverage). Google is now trying to close the gap, and have announced the inclusion of street/road data for the following countries:

Afghanistan, Aruba, Bahamas, Bahrain, Bangladesh, Belize, Bermuda, Brunei Darussalam, Bhutan, Bolivia, Cambodia, Chile, Columbia, Costa Rica, Cuba, Ecuador, El Salvador, Fiji, Guatemala, Haiti, Honduras, Indonesia, Iraq, Iran, Jordan, Kazakhstan, Kyrgyzstan, Kuwait, Laos, Lebanon, Mexico, Myanmar (Burma), Mongolia, Nepal, Nicaragua, Oman, Pakistan, Panama, Paraguay, Peru, Philippines, Papua New Guinea, Qatar, Saudi Arabia, Sri Lanka, Syria, Tajikistan, Turkmenistan, Timor-Leste, Uruguay, Uzbekistan, Venezuela, Vietnam, Yemen

It is great to see the market forces at play here - keeping the competition heated. However, my initial inspection of the data (that is to say, a purely ad hoc look at the first place I wanted to learn more about: the Mayan Riviera) suggested that Google hasn't really QAd this update. Here, for example, we see a major inaccuracy of road data in Mexico:

The 307 doesn't lurch into the ocean north of Playa Del Carmen (I know, I've driven down it).

It is often said that precision is the key metric in search - ought it not to be so in mapping as well?