SPEED VERSUS ACCURACY
BANTAY GOBYERNO SERIES 055
By Ike Señeres 02/12/2011
SPEED VERSUS ACCURACY
A concerned citizen asked me a layman’s question that needed a technical answer. He wanted to know what is more important in the automation of the COMELEC voting system. Is it speed or accuracy? I told him that that is precisely the challenge in the automation goal, to strike a good balance between speed and accuracy.
The equilibrium between speed and accuracy is needed because without it, the result could tilt towards one of the two extremes, and that could spell disaster. In either case, speed without accuracy or accuracy without speed would not be good for the system. By comparison however, accuracy would have a higher order of importance, because it would reflect the will of the people. Nonetheless, data derived from a slow process would still be questionable, because the people would suspect that the data might have been compromised in between.
Before the idea of automation came to the COMELEC, all votes were cast, counted, tabulated and transmitted manually. During the last election, the votes were counted and transmitted electronically, but the casting and tabulation stages were still done manually, at least from the technical standpoint. By comparison, the objective of speed was achieved, but some issues were still raised about the issue of accuracy.
There is a general impression that the casting of the votes in the last election was already automated. This notion is debatable, because the votes were still cast manually by using the Optical Mark Reader (OMR) technology. In using the OMR technology, the manual voting method simply changed from the use of hand written characters to the use of shaded dots, the latter representing optical marks. In a figurative sense, the second step of the voting was also manual, because the OMR ballots were manually fed into the Precinct Count Optical Scan (PCOS) machine.
There is also a general impression that the tabulation of the votes was automated, but this too is debatable. The reason for this is that the Compact Flash (CF) cards were manually removed from the PCOS machines so that these could be inserted into the CF card readers that were attached to personal computers (PCs) that were running on Windows Operating Systems (Windows O/S).
It could be said that the majority of the people were happy with the fact that the last election was able to deliver the expected results, even if the scope of the automation was only partial, so to speak. What is good however could be made better, more so because the law requires full automation, and not just partial. The law is the law, and we have to implement it, even if we are already happy with what we see.
When I chaired the COMELEC Modernization Committee twice in the past, I always encountered the issue of the disparity between what the law requires, and what our culture promotes or allows. The tolerance of partial automation is just one example, but there are more examples like the use of COMELEC identification cards as a requirement for the issuance of official ballots, and even as a control for the entry into the polling places, for security reasons.
What is the use of issuing COMELEC identification cards if these are not really required at the polling places? This is a good question that a reader asked me. A reader also asked me about the wisdom of using the Unified Multi-Purpose ID (UMID) as an alternative identification card. UMID is issued by the Government Service Insurance System (GSIS), but it is now being proposed as a unified card for all members of the Social Security System, PHILHEALTH and PAG-IBIG as well.
The issue of accuracy is directly related to the need for data integrity and data security. It is irresponsible for any organization to claim that their system is “hack-free”, because in theory all systems are prone to being hacked either partially or fully. This is precisely the objective of data security systems, to see to it that it could not be hacked quickly or easily, and to make it too costly for anyone to even try to hack it. The word to use is “super-secure”, not “hack-free”.
Conversely, there are technologies that would prevent a system from being fully hacked. This is a question of techno-economics; because the more money invested into the system to secure it would make it more costly to hack it. All told, the most secure data security system is one that would combine hardware encryption and software encryption technologies, not just one or the other. Crucial to this is the full legal ownership of all source codes and all passwords.
Go for the good life! Tune in to Universal Access 2 to 3pm Monday to Sunday in DZRJ 810 KHZ and to KA IKING LIVE! 6 to 7pm Thursdays in Global News Network (GNN), Channel 8 in Destiny Cable. Email iseneres@yahoo.com or send text to +639997333011for local cable listings. Visit www.senseneres.blogspot.com
By Ike Señeres 02/12/2011
SPEED VERSUS ACCURACY
A concerned citizen asked me a layman’s question that needed a technical answer. He wanted to know what is more important in the automation of the COMELEC voting system. Is it speed or accuracy? I told him that that is precisely the challenge in the automation goal, to strike a good balance between speed and accuracy.
The equilibrium between speed and accuracy is needed because without it, the result could tilt towards one of the two extremes, and that could spell disaster. In either case, speed without accuracy or accuracy without speed would not be good for the system. By comparison however, accuracy would have a higher order of importance, because it would reflect the will of the people. Nonetheless, data derived from a slow process would still be questionable, because the people would suspect that the data might have been compromised in between.
Before the idea of automation came to the COMELEC, all votes were cast, counted, tabulated and transmitted manually. During the last election, the votes were counted and transmitted electronically, but the casting and tabulation stages were still done manually, at least from the technical standpoint. By comparison, the objective of speed was achieved, but some issues were still raised about the issue of accuracy.
There is a general impression that the casting of the votes in the last election was already automated. This notion is debatable, because the votes were still cast manually by using the Optical Mark Reader (OMR) technology. In using the OMR technology, the manual voting method simply changed from the use of hand written characters to the use of shaded dots, the latter representing optical marks. In a figurative sense, the second step of the voting was also manual, because the OMR ballots were manually fed into the Precinct Count Optical Scan (PCOS) machine.
There is also a general impression that the tabulation of the votes was automated, but this too is debatable. The reason for this is that the Compact Flash (CF) cards were manually removed from the PCOS machines so that these could be inserted into the CF card readers that were attached to personal computers (PCs) that were running on Windows Operating Systems (Windows O/S).
It could be said that the majority of the people were happy with the fact that the last election was able to deliver the expected results, even if the scope of the automation was only partial, so to speak. What is good however could be made better, more so because the law requires full automation, and not just partial. The law is the law, and we have to implement it, even if we are already happy with what we see.
When I chaired the COMELEC Modernization Committee twice in the past, I always encountered the issue of the disparity between what the law requires, and what our culture promotes or allows. The tolerance of partial automation is just one example, but there are more examples like the use of COMELEC identification cards as a requirement for the issuance of official ballots, and even as a control for the entry into the polling places, for security reasons.
What is the use of issuing COMELEC identification cards if these are not really required at the polling places? This is a good question that a reader asked me. A reader also asked me about the wisdom of using the Unified Multi-Purpose ID (UMID) as an alternative identification card. UMID is issued by the Government Service Insurance System (GSIS), but it is now being proposed as a unified card for all members of the Social Security System, PHILHEALTH and PAG-IBIG as well.
The issue of accuracy is directly related to the need for data integrity and data security. It is irresponsible for any organization to claim that their system is “hack-free”, because in theory all systems are prone to being hacked either partially or fully. This is precisely the objective of data security systems, to see to it that it could not be hacked quickly or easily, and to make it too costly for anyone to even try to hack it. The word to use is “super-secure”, not “hack-free”.
Conversely, there are technologies that would prevent a system from being fully hacked. This is a question of techno-economics; because the more money invested into the system to secure it would make it more costly to hack it. All told, the most secure data security system is one that would combine hardware encryption and software encryption technologies, not just one or the other. Crucial to this is the full legal ownership of all source codes and all passwords.
Go for the good life! Tune in to Universal Access 2 to 3pm Monday to Sunday in DZRJ 810 KHZ and to KA IKING LIVE! 6 to 7pm Thursdays in Global News Network (GNN), Channel 8 in Destiny Cable. Email iseneres@yahoo.com or send text to +639997333011for local cable listings. Visit www.senseneres.blogspot.com
posted by Ike Señeres at 11:35 PM
0 Comments:
Post a Comment
<< Home