Tag Archives: semantic

Feb 17 2011

Instant Outline? Topicmarks Summarizes Complex Articles

Ho ho, now here is a COOL tool. A cool SEMANTIC tool. Particularly if you are lazy or reading comprehension isn’t your strong suit. MakeUseOf has a thorough review of Topicmarks, a web-based application that automates a summary of key terms and text in electronic form, or a “smart, interactive synopsis” of your electronic documents.

Open a free account, or sign in using Google or Yahoo. Topicmarks invites you to “open” your “personal knowledge space.” Then, upload a document (in a variety of formats, including Word, PDF, html, or plain text) to the Topicmarks site. Or use their bookmarklet to summarize any web page you happen to be visiting. Or email your document to the special email address provided by Topicmarks under the “profile” button.

Topicmarks’ engine then goes to work, and you will be notified by email when the process is complete. Under the button for “text knodes”, you will find your document summaries. Via tabs, you can get the overview, facts, summary, keywords, index, and properties. These tabs offer different ways to digest your information – either by quick facts, general overview, a deeper dive into particular keywords, or an index of all keywords. The information is linked across tabs – click on a keyword in the index and pull up the “facts” associated with that keword in the facts tab.

Your texts are stored forever, and it is currently free, although heavy users might see a cost for the service in the future.  Keep in mind that texts are public by default, and you can share your knodes with Facebook and Twitter, but you can make them private in the settings.

So, you research your topic (or you collect materials sent to you by others). Send them through Topicmarks. Read the synopsis. Check out some key terms. Does it look interesting? Check out the whole article? Does it look like dreck? Move on.

Now, it doesn’t take a rocket scientist, but it might take a lawyer, to see how Topicmarks might be of value to the legal profession. The fine folk at Topicmarks are mindful of this, including lawyers and paralegals on their list of who might benefit from their service. They also suggest the following uses for their awesome service:

  • Analyzing your own writings
  • Being up-to-date with the latest financial research
  • Building a knowledge base for a graduate thesis
  • Building a knowledge base for a master’s thesis
  • Checking current facts against past press releases
  • Checking doctoral theses
  • Discovering emerging patterns
  • Evaluating student papers quickly
  • Finding back quotes they remember having read somewhere
  • Finding inconsistencies in long reports
  • Getting the gist of subordinates’ presentations
  • Preparing a school project
  • Reading up quickly on industry analyses
  • Researching a first student paper
  • Sifting through annual reports
  • Sifting through legal cases
  • Staying abreast of white papers
  • Storing relevant legal precedents
  • Writing fiction abstracts y analyses
  • Researching a first student paper
  • Sifting through annual reports
  • Sifting through legal cases
  • Staying abreast of white papers
  • Storing relevant legal precedents
  • Writing fiction abstracts

Why do I love this? Well, it’s free (for now), it’s effective, its a serious efficiency tool, it’s web-based with social sharing, it’s research-and-writing-oriented, it’s uber-cool Semantic-powered. A clear winner on all counts. My mind is already coming up with new ways to play with it – I am thinking about finding the most complex securities case I can find in Google Scholar and running it through Topicmarks to see what I get. Check it out and turbo-charge your own electronic research and reading experience!

Advertisement
Aug 24 2010

Search Tools and Concept Searching

“Search” is familiar territory for lawyers – the term impliedly describes computer-aided discovery, evolved  from the longer, two-syllable term “research”, which, of course, includes old school book-based methods. But search itself is an evolving concept. We are far from the days of early Boolean search on a Walt II terminal. Furthermore, electronic search for lawyers now must adress the needs of both the legal researcher and the e-discovery expert.

I stumbled across an article today at Law.com’s Law Technology News entitled “What Lawyers Need To Know About Search Tools.” The article is written byMark Grossman and Terry Sweeney. Glancing only at the title, I thought it might be a *yawn* list of free on-line legal resources or *yawn* 1L explanation of how to form a search query. Not so. Written from the e-discovery perspective with the apparent intent of helping lawyers properly respond to discovery requests and comply with the Rules of Civil Procedure, the article is a decent critique of keyword searching as an effective means of information retrieval, particularly over vast stores of e-documents.  

By far, the best part of the article is the section entitled A VARIETY OF MODELS, which discusses variations on the theme of concept searching. The authors explain how taxonomies and ontologies (oooh, semantic!) can power-up search systems, retrieving far more relevant results than simple keyword search, particularly if beefed up with mathematical or statistical algorithms and/or combined with metadata. The authors concede the keyword and Boolean searching is still useful for small and rolling data sets. Nonetheless, it is clear that new search models are pushing our ability to quickly find and categorize information and changing the very definition of what constitutes a “reasonable” search.

The authors conclude that perfection is not the goal and reasonableness is, well, reasonable in e-discovery. My take away is a little different – the better our machine language / learning systems, tighter the algorithms and more consistent and deep the metadata, the closer we will get to perfection in search results. I, for one, can’t wait.

Dec 23 2009

Semantic Research For The Rest of Us

Popego
Image by magerleagues via Flickr

While stumbling through the reader tonight, I tripped upon an article by Dana Oshiro at ReadWriteWeb about a new semantic search engine called Meaningtool and a semantically-inclined feed generator called Popego.

These tools will help you cut down the noise and pull in the signal based on your interests and intended targets. After completing a profile, Popego provides you with semantic recommendations based on your on-line activity and social circles. Once you generate what Popego calls your “interest platform”, you can find more quickly relevant content and connect more readily with others based on your interests. Popego’s customized feeds can be widgetized and shared on your other sites and blogs.

Meaningtool will analyze any website, pull out the relevant terms and create categories based on these terms, generating a tag cloud. You can click on any of the categorizes and further refine and “tree” the resulting information. Meaningtool recognizes most of the popular Western languages. Meaningtool’s Category Manager allows you to “train” your semantic feed via relevant RSS feeds to get better results.

Meaningtool is currently being used by marketing firms to better target customers with advertising campaigns, publishers looking to improve their search engine ranking, and semantic Web developers.

Like my6sense, the iPhone tool that pulls the cream from your streams to the top, you need to train Popego and Meaningtool to get the most out of them. With the dizzyingly large amount of content on the Web, a little training goes a long way to bringing you want you want to see.

For laughs, I fed my own website’s URL into Meaningtool to see what would happen. After the engine finished thinking, it showed this graphic:

Scrolling down, the results expanded to a tag cloud and suggestions to improve my SEO (which desperately needs improving):

Finally, the bottom of the screen showed the most relevant terms according to the listed classifiers from services like the New York Times, LA Times, Reuters, and Digg:

For certain applications, I can see this being a very effective tool.

Check out the Meaningtool video here:

Popego’s demo video follows:

Related articles by Zemanta
Reblog this post [with Zemanta]
Dec 3 2009

“Dumbing” Down Search: Bing vs. Google

When looking for information on the Web, what search service do you turn to these days? Are you still a diehard Google fan and feel most comfortable with pages of popular links based on loosely associated keywords that can be tweaked by advance search form, which you can then follow to find more information and, ultimately, your “best” answer? Do you scan the source of those links on the Google results page to make your own assessments as to the potential veracity of the information excerpted? Do you seesaw back and forth between results page and linked sites until you are satisfied that you have the skinny?

Or, would you rather just get an answer and run? If you fall into this growing category of on-line searchers, then maybe Microsoft’s search upstart Bing is your answer.

Paul Boutin at VentureBeat offers the observation that Google’s search methodology is quickly becoming obsolete and that Bing is offering the better option for today’s searchers:. In Boutin’s own words “[i]n short, the people who use search engines today are nothing like the people who build them. Online, the normals have finally displaced the geeks.” His bullet points encapsulate the new search mentality:

  • Don’t give me a link to the answer. Just give me the answer.
  • Pictures are better than words.
  • I’m totally fine with getting search results from a Microsoft database of multimedia celebrity flash cards instead of from the entire Internet, if it tells me what I want to know

Bing meets these needs beautifully and has garnered ten percent of search traffic in the approximately 6 months it has been in business. From canned “cards” prepared by Bing-heads for frequently searched topics to answers pulled directly from Wikipedia, searches can easily “click and run” with their answers on Bing.

Google, concededly, requires a bit more effort. But Google offers the opportunity to view many different answers to a particular query and weigh the results based on the strength of the site from which the information is pulled.

I definitely see a place for both strategies in the search arena. But much like yesterday’s diatribe on the use of Wikipedia (never in a court room), searchers need to be cognizant of what they are getting and consider when the extra effort is necessary. Like Boutin, I sincerely hope that Google does not bow to the peer pressure and “dumb down” search. I still get a pitter patter in my heart when I see services like Wolfram/Alpha and the various semantic search tools providing yet another angle on the information. Because, as we all know, rarely is there one single true answer to any question, even the query: how do I get from Gloucester to Boston? There are still geeks out there like me who want to be shown the how and why of it.

You can call me the Norm Abrams of Search.

In a (barely) related note, I came across an interesting new bookmarking service this morning called Faviki. Like Delicious, it allows you to collect and tag your bookmarked sites. Unlike Delicious, it offers suggested tags based on semantic overlay. Unfortunately, the semantic overlay is from Wikipedia, as this still remains the largest repository of common information on the Web. Nonetheless, it is an interesting application of semantic technology and may be worth a try. Particularly since it now allows for importing your huge library of Delicious bookmarks.

Dec 3 2009

"Dumbing" Down Search: Bing vs. Google

When looking for information on the Web, what search service do you turn to these days? Are you still a diehard Google fan and feel most comfortable with pages of popular links based on loosely associated keywords that can be tweaked by advance search form, which you can then follow to find more information and, ultimately, your “best” answer? Do you scan the source of those links on the Google results page to make your own assessments as to the potential veracity of the information excerpted? Do you seesaw back and forth between results page and linked sites until you are satisfied that you have the skinny?

Or, would you rather just get an answer and run? If you fall into this growing category of on-line searchers, then maybe Microsoft’s search upstart Bing is your answer.

Paul Boutin at VentureBeat offers the observation that Google’s search methodology is quickly becoming obsolete and that Bing is offering the better option for today’s searchers:. In Boutin’s own words “[i]n short, the people who use search engines today are nothing like the people who build them. Online, the normals have finally displaced the geeks.” His bullet points encapsulate the new search mentality:

  • Don’t give me a link to the answer. Just give me the answer.
  • Pictures are better than words.
  • I’m totally fine with getting search results from a Microsoft database of multimedia celebrity flash cards instead of from the entire Internet, if it tells me what I want to know

Bing meets these needs beautifully and has garnered ten percent of search traffic in the approximately 6 months it has been in business. From canned “cards” prepared by Bing-heads for frequently searched topics to answers pulled directly from Wikipedia, searches can easily “click and run” with their answers on Bing.

Google, concededly, requires a bit more effort. But Google offers the opportunity to view many different answers to a particular query and weigh the results based on the strength of the site from which the information is pulled.

I definitely see a place for both strategies in the search arena. But much like yesterday’s diatribe on the use of Wikipedia (never in a court room), searchers need to be cognizant of what they are getting and consider when the extra effort is necessary. Like Boutin, I sincerely hope that Google does not bow to the peer pressure and “dumb down” search. I still get a pitter patter in my heart when I see services like Wolfram/Alpha and the various semantic search tools providing yet another angle on the information. Because, as we all know, rarely is there one single true answer to any question, even the query: how do I get from Gloucester to Boston? There are still geeks out there like me who want to be shown the how and why of it.

You can call me the Norm Abrams of Search.

In a (barely) related note, I came across an interesting new bookmarking service this morning called Faviki. Like Delicious, it allows you to collect and tag your bookmarked sites. Unlike Delicious, it offers suggested tags based on semantic overlay. Unfortunately, the semantic overlay is from Wikipedia, as this still remains the largest repository of common information on the Web. Nonetheless, it is an interesting application of semantic technology and may be worth a try. Particularly since it now allows for importing your huge library of Delicious bookmarks.

Dec 1 2009

Examining the Supreme Court

Interested in the United States Supreme Court? Check out the Supreme Court Database, a broad-based collection of information not necessarily targeted at lawyers. Currently, the store holds over 200 pieces of information on all cases decided between the 1953 and 2008 terms. It is a work in progress and new information is being added all the time. The proponents come from a variety of universities and colleges and a range of backgrounds.

In a sense, the database seeks to impose a structure on the pertinent data within and surrounding each decision. There is an analysis tool that allows for searching across the database using concepts and keywords. Hit the link to the codebook to view the list of variables culled from the cases, searchable using the analysis tool – it really is quite impressive.

If such analysis can be applied to the Supreme Court’s decisions, why not the lower court cases and statutory law? Sure it might take a while, but this effort could open the door to a semantic treatment of legal resources. Maybe we could even push legal research out of the approximately 15-year slump in which it is been languishing. Are you listening, Google?

Jun 12 2009

Wolfram Alpha – Already Getting Bigger & Better

Less than a month old, the newborn Alpha is already celebrating by giving a gift back to its users: upgrades! Resource Shelf (and lots of others) report on the upgrades here. Resource Shelf quotes the changes reported on the Wolfram Alpha blog, and I requote them here for convenience:

Additional linguistic forms for many types of data and questions

More comparisons of composite properties (e.g. “US military vs. UK”)

Combined time series plots of different quantities (e.g. “germany gdp vs population”)

More complete handling of government positions (e.g. “chancellor”, etc.)

Updates to country borders for India, China, Slovenia, Croatia, and others

Updates to naming for certain politically sensitive countries and regions

Additional subcountry regions (e.g. “Wales”); many more to come

Additional support for current and past fractional timezones (e.g. “Iran time”)

City-by-city handling of U.S. states with multiple timezones

Updates to certain European currencies (e.g. for “Cyprus” and “Slovakia”)

Some additional historical events; many more to come

Additional probability computations for cards and coins (e.g. “2 or 3 aces”)

Additional output for partitions of integers (e.g. “partitions of 47″)

Implicit handling of geometric figure properties (e.g. “ellipse with area 6 and major axis 2″)

Additional support for Mathematica 3D graphics syntax

Additional support for stock prices with explicit dates

Support for planet-to-planet distances and “nearest planet”, etc.

Extra information when comparing incompatible units (e.g. “ergs vs. newtons”)

Improved linguistic handling for many foods (e.g. “love apple”)

More mountains added, especially in Australia

Support for many less-common given names (e.g. “zebulon”)

More “self-aware” questions answered (e.g. “how old are you”)

More consistent handling of sidebar links to Wikipedia, etc.

Happy Birthday to us!

[Caption]

Reblog this post [with Zemanta]
Jun 11 2009

Can Legal Information Be Semanticized?

You may have noticed that it has been unusually quite in the Studio over the past week. Extreme travel and excessive work have taken their toll on my blogging output. However, it is time, once again, to buckle down and plumb the depths of law, technology and cutting-edge research and writing tools.

Absence apparently makes the heart grow fonder, as my reader greeted me with an interesting article over at Law.com by Dr. Adam Zachary Wyner on legal ontologies and how they may “spin a semantic web”.  I have heard it said before that legal information is too broad and deep and resistant to organization to be put under the semantic knife, or more aptly “blanket.”  I was heartened to see that Dr. Wyner disagreed (sort of) with this position.

First, I commend Dr. Wyner on providing an excellent explanation of what ontologies, knowledge management and taxonomies are and how they interrelate. I learned a few new concepts reading through his summary.

Next, the problem for those of us anxiously awaiting semantic treatment for case and statutory law is indeed that there is little consistency in how these materials are presented, the wide variety of search need to be satisfied and the sheer volume of material produced on a yearly basis. Dr. Wyner sees this conundrum as well. Coupled with the fact that most text marking, the process by which information is overlaid so that a computer can “read” and calcaulate an answer to a semantic search, would have to be performed by humans in the legal arena, it appears that the desired result may be somewhat of an impossible dream.

However, Dr. Wyner correctly points out that there are some categories of information that are clearly defined across cases and statutes that could certainly be marked for organization. For example, in the case law context, case headings, party information, result, even statements of issues can be tagged, organized and converted to computer-readable form. Dr. Wyner even suggests means for treating the information: through Semantic MediaWikis, as a learning tool for researchers and law students, and through treatment by large scale publishers and government agencies and courts prior to publication. Dr. Wyner sounds as enthusiastic as I feel about the possibilities of incorporating this layer of information and what such treatment could bring to the legal researcher’s table.

I am relunctant to get too excited about the Big Two jumping onto the semantic bandwagon and spearheading the effort to “semanticize” the vast amounts of legal information in their stables. After all, where is their incentive? Nonetheless, I can’t help but smile a bit at Dr. Wyner’s excited undertone and the thought that bringing cases and statutes into the 21st century may not be as impossible a dream as Don Quixote tilting at windmills.

Reblog this post [with Zemanta]