Don Dew of advanced document recognition ISV Parascript recently authored an article on the benefits of intelligent character recognition (IDR) technology that he has asked us to share with you. It discusses how ICR technology is underutilized on documents that include handprint and even cursive writing.
Here's an excerpt from Dew's article, "From name, address, social security number, phone number,
or any other unconstrained or cursive information entered on a form,
advanced ICR solutions can capture this data with a high degree of
accuracy and make it available for use within the organization. Based on research performed by AIIM and
Parascript last year, only 6% of organizations are automating this level
of recognition. At the same time, survey participants estimated that
they would achieve a considerable level of productivity savings if they
were able to automate the recognition of hand-written text."
Here's the link to Dew's complete article on our DIR Web site.
Just for some background, Parascript develops advanced recognition technology that does both OCR and ICR. Historically, it has been best known for its cursive recognition, which is utilized by the USPS in their envelope and parcel sorting operations. Parascript also offers a document recognition-centric toolkit that includes both OCR and ICR. Parascript recently released a new version of this SDK, as well as a new version of its FormXtra Capture application, which it is moving through a recently revamped reseller program.
Showing posts with label Forms Processing. Show all posts
Showing posts with label Forms Processing. Show all posts
Monday, May 06, 2013
Monday, March 04, 2013
OCR/ICR Survey Highlights Infographic
We originally highlighted the results of this AIIM survey, (sponsored by Parascript) when it first came out last summer. (Download the entire report here).
Some interesting outtakes, some of which have been highlighted below in an outstanding infographic:
Some interesting outtakes, some of which have been highlighted below in an outstanding infographic:
- 55% of respondents who were scanning documents were key entering data
- Only 32% were using OCR with ICR and cursive recognition utilization significantly lower
Wednesday, August 22, 2012
Forrester Capture Report
Has anyone seen this new Forrester Wave report on Multi-Channel Capture? Kofax, which is absolutely the leader, top and to the right with EMC Captiva slightly behind, is making the report available if you register. There is also an AIIM Webinar scheduled for next week.
Anyhow, the report has caused some controversy in the industry. Most obviously, there are only nine vendors ranked, when seriously I think there are about 100 competing in the document capture software space - not to mention the "multi-channel" capture market, which would seem to bring even more ISVs into the mix.
I'll start off by saying that I like the term "multi-channel capture" as a moniker that identifies where the industry is headed. Automating the capture of data from paper documents, while greatly increasing efficiencies, is not enough anymore as paper use declines and new forms of electronic input emerge. I also like Forrester's stressing the analytics is going to become an increasingly important part of this market going forward. I'm not sure they utilizing the same framing of "analytics" that I would, but I think their concepts in this area are strong at least.
On the flip side, ranking only eight vendors seems like a great disservice. I mean they left out clear market leaders like Nuance and ReadSoft, while ranking their competitors like NSi, IteSoft, Brainware, and TIS. Not sure of the rhyme or reason behind the vendor choices although Forrester attempts to link it to customer demand. Some speculation is that Forrester's background is in ECM and not data capture, so maybe their choices were influenced by that. Either way, I think the number of vendors included comes up way short.
I'll have some more details in an upcoming premium issue, but as a word of warning, here's what one ISV (whose company was ranked) said to me regarding the report: "Something like this is supposed to help clear up confusion in the market for end users. This report only adds to it."
Anyhow, the report has caused some controversy in the industry. Most obviously, there are only nine vendors ranked, when seriously I think there are about 100 competing in the document capture software space - not to mention the "multi-channel" capture market, which would seem to bring even more ISVs into the mix.
I'll start off by saying that I like the term "multi-channel capture" as a moniker that identifies where the industry is headed. Automating the capture of data from paper documents, while greatly increasing efficiencies, is not enough anymore as paper use declines and new forms of electronic input emerge. I also like Forrester's stressing the analytics is going to become an increasingly important part of this market going forward. I'm not sure they utilizing the same framing of "analytics" that I would, but I think their concepts in this area are strong at least.
On the flip side, ranking only eight vendors seems like a great disservice. I mean they left out clear market leaders like Nuance and ReadSoft, while ranking their competitors like NSi, IteSoft, Brainware, and TIS. Not sure of the rhyme or reason behind the vendor choices although Forrester attempts to link it to customer demand. Some speculation is that Forrester's background is in ECM and not data capture, so maybe their choices were influenced by that. Either way, I think the number of vendors included comes up way short.
I'll have some more details in an upcoming premium issue, but as a word of warning, here's what one ISV (whose company was ranked) said to me regarding the report: "Something like this is supposed to help clear up confusion in the market for end users. This report only adds to it."
Wednesday, August 15, 2012
Thoughts on Parascript/AIIM Forms Processing Study & White Paper
Lot of interesting information in a recent study conducted by AIIM and sponsored by Parascript.
Parascript develops a slew of recognition technology including handprint and cursive recognition. Not surprisingly, a follow-up article written by Parascript's Don Dew highlights some of the shortcomings in adoption of handprint/cursive recognition. According to Dew, "In most organizations, hand-written fields are prevalent on a significant number of forms. 42% of respondents indicated they have hand-written data fields on half or more of their forms. In addition to being prevalent, these hand-written forms are also important to the efficiency of the business process. 40% of respondents say they are quite important; 20% say they play a key role."
"However, many organizations are not taking advantage of this information. 88% of respondents say they scan forms, but only 32% say they perform text recognition to automatically make that data readily available for use in their organizations. The majority of respondents (55%) report they scan images and manually re-key the data as part of their workflow."
Of course, this is where Parascript's technology could come in, or, the crowdsourcing data-entry solutions from companies like virtualsolutions and Captricity, which were featured in our last premium issue. Some combination of the two may actually form the most efficient solution.
Another interesting point made in the study is that users cited a multitude of forms as the number one reason that they are not using forms processing technology - in other words, they feel the templates are too hard to set up. This should be interesting news to companies like ITyX, a German artificial intelligence vendor that DIR was recently introduced to.
Anyhow, there is a lot of interesting stuff in this white paper about forms processing adoption, what end users are implementing it for, and why they are not in certain areas. The bottom line to me seems to be that parochial/departmental management of many forms capture operations prevents users from looking at the top tier capture automation solutions out there. They just don't have the bandwidth to consider the cutting edge technology that is most often included in enterprise capture applications. SaaS/Cloud services may prove to be the way around this.
Parascript develops a slew of recognition technology including handprint and cursive recognition. Not surprisingly, a follow-up article written by Parascript's Don Dew highlights some of the shortcomings in adoption of handprint/cursive recognition. According to Dew, "In most organizations, hand-written fields are prevalent on a significant number of forms. 42% of respondents indicated they have hand-written data fields on half or more of their forms. In addition to being prevalent, these hand-written forms are also important to the efficiency of the business process. 40% of respondents say they are quite important; 20% say they play a key role."
"However, many organizations are not taking advantage of this information. 88% of respondents say they scan forms, but only 32% say they perform text recognition to automatically make that data readily available for use in their organizations. The majority of respondents (55%) report they scan images and manually re-key the data as part of their workflow."
Of course, this is where Parascript's technology could come in, or, the crowdsourcing data-entry solutions from companies like virtualsolutions and Captricity, which were featured in our last premium issue. Some combination of the two may actually form the most efficient solution.
Another interesting point made in the study is that users cited a multitude of forms as the number one reason that they are not using forms processing technology - in other words, they feel the templates are too hard to set up. This should be interesting news to companies like ITyX, a German artificial intelligence vendor that DIR was recently introduced to.
Anyhow, there is a lot of interesting stuff in this white paper about forms processing adoption, what end users are implementing it for, and why they are not in certain areas. The bottom line to me seems to be that parochial/departmental management of many forms capture operations prevents users from looking at the top tier capture automation solutions out there. They just don't have the bandwidth to consider the cutting edge technology that is most often included in enterprise capture applications. SaaS/Cloud services may prove to be the way around this.
Monday, January 31, 2011
OCR Accuracy Benefits
One of the often under-rated benefits of automated data capture technology is the accuracy improvement it can bring to a process. I was just typing a colleague's business card info. into Outlook (yes, I still sometimes use Outlook), and I accidentally typed the fax number into the field for cell phone. Luckily I caught it and saved myself the pain of hearing a loud screeching sound when I thought I was dialing a mobile phone number. As I typed in the correct number, I realized that would not have happend had a I scanned the card and used my OCR-enabled business card software, which I typically use for batch capture only.
So, yes, while we usually like to stress the productivity benefits - less manual keying - of OCR applications, there are certainly accuracy benefits as well, that can potentially further contribute to the ROI - how much depends on the value of your downstream data. Which brings me to a second instance of inaccurate data entry that I ran into this weekend: My wife had recently purchased a memorial service (I don't know if purchased is the right word, but...) for her father-in-law in our family's name. Well, wasn't it quite a surprise, when I received the church bulletin this week and saw the service was being held in memory of me! Whoops. The flowers have certainly brightened up the house though.
So, yes, while we usually like to stress the productivity benefits - less manual keying - of OCR applications, there are certainly accuracy benefits as well, that can potentially further contribute to the ROI - how much depends on the value of your downstream data. Which brings me to a second instance of inaccurate data entry that I ran into this weekend: My wife had recently purchased a memorial service (I don't know if purchased is the right word, but...) for her father-in-law in our family's name. Well, wasn't it quite a surprise, when I received the church bulletin this week and saw the service was being held in memory of me! Whoops. The flowers have certainly brightened up the house though.
Wednesday, March 07, 2007
Datacap Parascript form partnership
Automated data capture specialist Datacap, out of Tarrytown, NY, has officially announced a partnership with recognition specialist Parascript. Parascript, which was originally founded in Russia as part of the Glasnost initiative, is now based in Boulder, CO. It's claim to fame and biggest source of revenue is a contract with the USPS for reading addresses on envelopes. Parascript is noted for its ability to recognize cursive writing and has done a couple of installations with Datacap already involving hand-filled order forms for the Baltimore Sun and TV Guide. Parascript appears to be moving forward after a proposed merger with pubically traded Mitek fell through. Both Mitek and Parascript have strong technology in the check recognition market.
Subscribe to:
Posts (Atom)