The Semantic Desktop? SDS Brings Semantics To Excel
When you hear the word "semantic" you likely think of the semantic web - the supposed next iteration of the World Wide Web that features structured data and specific protocols that aim to bring about an "intelligent" web. But the concept of semantics doesn't necessarily apply just to the web - it can apply to other things as well, like your desktop...or even your Excel spreadsheets, according to Ian Goldsmid, founder of Semantic Business Intelligence, whose new app, SDS, brings a semantic system to spreadsheets.
Semantic Spreadsheets
The problem with spreadsheets that their system is trying to address has to do with those who need to derive data from multiple spreadsheets (two or more). Although it's easy enough to perform sorts, build macros, and create formulas within one spreadsheet, when needing to compare values in multiple spreadsheets the process becomes more difficult.
The company's app, The Semantic Discovery System for Excel, or just SDS for short, will look for similar columns or rows between the sheets and then "semantically" connects them. They don't appear to just be throwing that term around either - the app uses the same W3C Semantic Web technologies (RDF, OWL, SPARQL) to help you capture "meaning, intelligence, and knowledge" from the data saved in your spreadsheets.
Do We Need Semantic Desktop Apps?
Does SDS solve a business problem that is not yet being addressed through current technologies? In my experience, the short answer to this question is "no." (But wait, there's more...)
Typically, when a business has need of comparing and analyzing large amounts of data, the solution is to turn to a database product that can then be queried and from which custom reports can be pulled. And a business doesn't need to spend a lot of money on a robust solution to do so - even a smaller business can create a database by using inexpensive desktop software.
However, the difference between using a database technology and "semantically connecting" some spreadsheets comes down to for whom this product is being built. In the past, databases and other business intelligence apps were built as if the creators knew that the only person using them would be an I.T. guy or gal. SDS, instead, aims to satisfy the needs of the non-technical end user.
Is this another example of tech populism at work? It certainly looks like it. Yet, in this case their market is small - a non-technical user who's also a power user with Excel? There's usually some overlap there. Not to mention, by the time you've achieved "power user" status, you've often also figured out how to do more complicated things in Excel...like, say, formulas that work across spreadsheets, for example - the very pain points this app is trying to address.
Still, it's an interesting concept to think of taking the semantic web capabilities and integrating them into everyday programs to add a layer of intelligence to these programs as well. Done correctly, it could improve the capabilities of our favorite software apps without making the programs overly complex, which is what typically happens when you add more features.
What do you think? Is the Semantic Desktop (that is, semantically-enabled desktop apps) right around the corner? Or is this product and those like it too niche to find an audience? Let us know what you think in the comments.
Share
Comments
Subscribe to comments for this post OR Subscribe to comments for all ReadWriteWeb posts
Hm. With over a decade experience in this kind of data magic, I would note that if you are able to understand their system, you should be capable enough of understanding how to structure your excel files smart and efficient in order to work with them.
This tool might still be useful for advanced stuff, but starting of the video telling somebody that Excel is bad at connecting two simple tables like that is just ... well: not encouraging to believe what they are doing.
Now back to data as you described it: Either it is in the hand of one such knowlegable person or a few ones, with strict rules and processes in place who works on what or you do have bigger guns, like olap tools when working with most common spreadsheet data.
Either way: 95% of all users of office tools have really no clue how to use them beyond a standard set of maybe 10 features. I look at their presentation above and say "shiny", but normal users? I highly doubt that.
That is interesting,
At thoughtexpress.com we have created a semantic structure that in some way can be used as a super structured Excel (lets call it semantic Excel in its simplest form). In it is much easier to express all the rules of calculations, BPM etc then in Excel. Behind this sits our semantic computing cloud that allows to run generic enterprise or on a small scale one's models. So rather than discover semantics in Excel (bad forms of expression) create one where one can express it better. Nothing more ugly than replicating cells in Excel to deal with multiplicity.
Finally, semantic human interface creates dynamic user interaction based on knowledge itself.
All this semantic stuff is definitevely on the right track. Knowlegde/semantics should drive everything.
Pawel Lubczonok
This could also be automatic schema creation, even automatic conversion from databases to semantically linked data. And applied in that way it could be genuinely useful to many people.
Howeverfrom certain angle this seems like a solution looking for a problem. But nevertheless, I think that trying to build something new and more automatic will on the long run bring new qualities to web and computing. We just have to keep on trying and being creative with "the graph".
Oh, and it is interesting to see how smaller semantic web shops are trying to use W3C semantic standards as much as possible while high profile ones (Freebase/Metaweb, Twine, ...) are internally rolling their own technologies for dealing with linked data. I am also finding W3C stuff a bit clumsy to work with internally - but it is a nice interchange standard.
Andraz Tori, Zemanta
Some quick thoughts on who the target for such tools might really be.
In Life Science Drug research the Excel problem we face is different to that perhaps seen in mainstream B.I. Here, we have all of the 'hardware' (machines for sequencing Gene data etc) debouching their large volumes of data to ... Excel. Everyone seems to use Excel as the data 'dump' standard, so large volumes of Gene, Protein, Compound etc data is produced every day and the issue is having to make sense of the separate worksheets by in some way gluing them together to become a poor persons database. Today much cutting and pasting is needed before we can get to one neat coherent (very wide) spreadsheet that we can even think of using to start asking questions. So the problem is one of 'Excel as faux database' - I notice the US GOvt EPA indicate a similar problem with 'how to make sense of/consolidate a profusion of spreadsheets - http://articles.directorym.net/How_Semantics_Can_Revolutionize_Spreadsheets-a921265.html
The company I work for, InetSoft, provides these capabilities in an easier way, and all on the web.
Check out our flash demo.
We coined the term Data Mashup to describe this dynamic user-defined combination of various sources.