Sunny’s blog posting reminded me of the effort for ‘Image-based’ search by using mobile phone – Vodafone and Nokia. Internet service companies are not the only ones who try to find out an effective and accurate way for ‘Image’ search. In order to provide better service to their customers and to invent their killer application, mobile telecommunication incumbents are also striving for this area.

• Vodafone‘ image-based mobile search ‘Otello’
  : At CeBIT 2008, Vodafone announced that they’ll be trialling their Otello image-based search technology in Germany. In other words, handset owners just snap a picture of anything — a landmark, DVD case, unidentified flying object, etc. — and Otello then “returns information relevant to the picture to the mobile phone.

• Nokia was also planning a semantic visual search engine, which makes plenty of sense given their push in high-quality cameras for their mobile phones.
  : The visual search engine uses three process levels to extract semantic information from an image.1) When analyzing the images, at first they are converted into a plurality of candidate low-level
  features (like shape, color and texture strength) and these features are extracted locally around salient points of the image.

  2) Then a supervised learning approach is used to select prominent low-level features from the plurality of candidate low-level features. The prominent low-level features are associated with predefined object categories, that describe generic objects (e.g., cars, planes, etc.); parts of a person’s body (e.g., faces), geographical landmarks (e.g., mountains, trees. etc.), or other items.

• 3) When a new item is to be categorized, the target item is converted into a plurality of multi-scale local features and then each local feature is matched with the prominent low-level features using a probabilistic model. So, if the target item has a face, then this feature will be matched accordingly to the other items having a face and the item will be categorized.

Besides these two players, even though I couldn’t find a success story so far, numerous mobile-related companies (including KDDI, iPhone) are working on this huge project continuously. However, this technology could one day be used to search for more information on famous places, DVDs, toys and so on by simply taking a picture with your handset while walking on the street.

BBC News is reporting that YouTube has removed some videos from its site that it judged to glorify the Columbine school shooters, which left me wondering what one does when one expunges “undesirable” data from a collection. Assuming the expunging is justified, do you keep the reference information so you have a record of having had the thing around (and thereby make yourself better able to detect its reappearance)? Do you expunge the thing from the entire database? It seems good general practice to have a place where one can keep old records that no longer point to something retrievable. Would it be wise to allow people to search and find that an item had been intentionally removed, to save them the trouble of searching and searching for it? Or would it be ethically questionable to have even just the record available, since it could give people the idea to seek it elsewhere or create copycat works? I’m guessing the videos will appear elsewhere on the net, and there is little anyone can do to keep them out of public view, but keeping them off popular sites could effectively marginalize them. I’m thinking the benefit of keeping something truly nasty beyond the view of the “tell me something about …” searcher outweighs the benefit of explaining the removal to the “I want this exact document” searcher.

Jeff Atwood has a post today about stopwords (as we discussed in class yesterday).

He shows the default lists of stopwords that ship with Microsoft SQL Server and Oracle, which are interesting to see, and posts some interesting numbers on frequency of words on the web.  He finds, as we might expect, that many of the most frequent words aren’t normally considered stop words (information, website, download, internet, home, email).  He also links to an interesting Google patent on analyzing when to ignore stop words and when not to.

Again, commenters add to the blog: it looks like Tim Bray did a similar analysis in 2003.  Both note that Google handles searches for “to be or not to be” correctly (though it sounds like the behavior today is better than the behavior in 2003).

I think it bears repeating that the commonness of these words doesn’t seem like a good reason to drop them from indices or search queries.  A word that appears in every document might be useless, but if I can halve the result set with a single word (I’m looking for an email address, say), then the relative frequency of the word “email” doesn’t seem to hurt me much.  Removing stopwords that are unlikely to have semantic value (articles and conjunctions, say) makes more sense to me.

The European Union just repealed the laws they had in place banning the sale of abnormally shaped vegetables.

Pretty weird, but even crazier is the way they classify and standardize definitions of deformed produce:

The European Union is well known for its detailed regulations on appropriate shapes and sizes for agricultural items. Commission Regulation (EC) 2257/94, for example, states that bananas sold in Europe must be “free from malformation or abnormal curvature,” though Class 1 bananas can have “slight defects of shape,” and Class 2 bananas can have full “defects of shape.” Bananas were not covered in Wednesday’s ruling, so for now, these standards remain.

Different classes depending on how many “defects in shape.”  nice.

On the recent All Tech Considered Monday segment of NPR’s afternoon program All Things Considered, a couple new cell phone applications were featured, first one being Shop Savvy which allows the phone to scan the barcode of anything from a book to a CD to a bottle of Snapple and outputs the cheapest price on the product as well as reviews from people who have purchased it and the second one being Shazam which reads the digital fingerprint of a song on the radio you are unable to recognize, tags the music and tells you what the album is along with where you can purchase it. Since it recognizes the digital fingerprint, it is unable to recognize human humming or singing and extract information from that…yet.

As we discussed during Monday’s discussion section, image, audio and visual recognition are growing increasingly powerful and user base for available tools will likely expand as technology continues to improve.

http://www.pressgazette.co.uk/story.asp?storyCode=41225&sectioncode=7

I came across this a bit old story about a project undertaken by Tim Berners lee. The project aims at creating a system to feed important metadata like journalist’s profile, way the news story was created etc. to create a rich metadata set which will provide more credibility to the news story and also provide more retrieval options.

It also aims to standardize the way this information can be embedded in the news stories to help news aggregators provide us with more accurate and meaningful news. 

“They can be buried anywhere – the first or second paragraph, the beginning of the story, or even the end,” he said. “It just seemed incredible that of all the basic information you might want to know about a story, even such basic things as who wrote it and for who, is extremely hard to get at the moment.”

http://scripts.mit.edu/~cci/HCI/index.php?title=Main_Page

Here is the discussion about the redesign of google’s advanced search page (thanks to Mano Marks for initially pointing this out). Dan Russell, the googler who posted it, describes some of the thinking that went into the changes, from the version we saw in class to the new and improved version. I think the visualization of the query string is a nice touch. To contrast the experience with other search engines: Yahoo! (Doesn’t seem to have advanced search, just suggestions, thanks for the correction ryan), Ask.com (reminiscent of google’s old style), and Microsoft’s Live.com (Live’s operates iteratively so the query is narrowed click by click, it ends up being functionally similar but a very different user experience), Dogpile (Pretty standard, but doesn’t take you to a different page).  Cuil, the upstart that got a lot of press a while back, doesn’t have any advanced search I could find.

In thinking about my personal use of advanced search, I tend to use it when I know a particular thing exists and I can use limiters (like filetype: or site:) to increase the specificity of my query and reduce the number of results without risking that I will exclude something unintentionally which may have been useful. So assuming that even a fraction of the 67% of people who in 2000 reported that they had been frustrated by a web search, (from the Resnick and Vaughn reading) can advanced search be of assistance? I’m guessing that it probably can, although it is hard to imagine a large portion of the user base taking an interest in compund boolean queries when they have the simple “white hole” as their alternative.

Google’s advanced search operators are explained on their tips page.

Searching for images on the Web has traditionally been more complicated than text search – for instance, a Google image search for “tiger” not only yields images of tigers, but also returns images of Tiger Woods, tiger sharks and many others that are ‘related’ to the text in the query string. This is because contemporary search engines look for images using any ‘text’ linked to images rather than the ‘content’ of the picture itself.  In an effort to improve the recall of image searches, folks from UC San Diego are working on a search engine that works differently – one that analyzes the image itself. “You might finally find all those unlabeled pictures of your kids playing soccer that are on your computer somewhere,” says Nuno Vasconcelos, a professor of electrical engineering at the UCSD Jacobs School of Engineering. They claim that their Supervised Multiclass Labeling System “may be folded into next-generation image search engines for the Internet; and in the shorter term, could be used to annotate and search commercial and private image collections.”

What is Supervised Multiclass Labeling System anyway?

Supervised refers to the fact that the users train the image labeling system to identify classes of objects, such as “tigers,” “mountains” and “blossoms,” by exposing the system to many different pictures of tigers, mountains and blossoms. The supervised approach allows the system to differentiate between similar visual concepts – such as polar bears and grizzly bears. In contrast, “unsupervised” approaches to the same technical challenges do not permit such fine-grained distinctions. “Multiclass” means that the training process can be repeated for many visual concepts. The same system can be trained to identify lions, tigers, trees, cars, rivers, mountains, sky or any concrete object. This is in contrast to systems that can answer just one question at a time, such as “Is there a horse in this picture?” (Abstract concepts like “happiness” are currently beyond the reach of the new system, however.) “Labeling” refers to the process of linking specific features within images directly to words that describe these features.

While the idea of searching images by their ‘content’ is indeed promising, there are some questions that still need to be answered. To what extent does the system’s efficiency depend on the sample of images used for training?  What is the impact of variations in the quality of photos on the algorithm’s performance? What big a role will these play in affecting the user’s supposedly improved search experience? Finally, do we foresee an extension of the algorithm to determine abstract concepts in the images too? Indeed, these are interesting areas to explore; nevertheless, the SML seems to be a significant step towards better image retrieval mechanisms.

Read more about the SML at http://www.jacobsschool.ucsd.edu/news/news_releases/release.sfe?id=650

A search engine that lets users benefit from their social network to improve search results:

delver.com

From the about us page:

What is Delver?

Delver is an intelligent social search engine that enables you to find, experience and benefit from the wealth of information created and referenced by your social world. Our mission is to empower you to easily discover and benefit from the collective wisdom of your social world. Your circle of friends and extended network are increasingly creating and sharing useful information and media online through: blogs, videos, reviews, articles, websites, music… and the list is only growing. By indexing all that shared knowledge, media, opinions, and activities, we can deliver search results that are truly relevant to you.