Διαχείριση συναλλαγών σε ένα σύστημα κτηματολογικών πληροφοριών. Ο αλγόριθμος FoCHAL

Abstract

Nowadays, Geoinformation Technologies are correlated with commercial software, thus ready solutions, which can satisfy in a great level the operational demands of Organizations administrating spatial data with Interoperability. Via the current GIS technology, a very large spatial database of an Organization in a multi-user work environment, allows limited functionality both in common editing of database in general and in common records more specifically. This common editing in the classic GIS industry is permitted only on one-level of cooperation among the users, nonetheless, in a way that creates «conflicts» with the editing of the same record; conflicts that have to be solved by a third user. This phenomenon is most frequent in Organizations operating with Interoperability, which are in need of inside and outside apportionment of information, dealing with enormous spatial datasets in long-term transactions. This doctoral thesis examines modern spatial databases and analyses their application as susceptible and autonomous GIS, ruled by the Principle of Spatial Interoperability and based on international standards. It also examines the phenomena in those multi used spatial databases and evolves the creation of an algorithm: the «FoCHAL» (Free of Conflicts Hierarchic - Altruistic Locking) algorithm. By FoCHAL algorithm, different users can edit simultaneously common records, preserving the database serialization intact with a proper and fully automated conflicts resolving mechanism. This new Geoinformation Technology is applied on the operational model of the Greek Cadastre Spatial Database, as an open and interoperated GIS, where exactly the developmental character of the Cadastre System is explicitly emerged, a fact that implies huge operating, political, social and financial benefits. Above all, the most important benefit is the preservation of serialization of the spatial database and the achievement of the «On-Line» operation. This operation allows the development of orthological SubLIS, a fact that stems from the spatial interoperability and allows the achievement of e-Government. The last decade, modern Geoinformation science has been established through remarkable efforts by software companies, evolving into what we all know today as a classic Land Information System solution. The focus of such a solution is the classic GIS software and its familiar user interface. Users and designers of LIS systems usually prefer a familiar interface and functionality of a commercial GIS. Such a solution ignores an important event: the IT (Information Technology) evolution, especially in Relational Data Base Management Systems (RDBMS) GIS functionality. The last years, these systems have been dramatically improved in spatial data viewing, management and analysis, with very remarkable results in operation capabilities, which are ignored by LIS and GIS designers. As a result, both LIS and GIS applications of modern Organizations that are using geoinformation technologies can be judged as "old fashioned", even incomplete, because they ignore this recent technological evolution. In the last past months, a common subject of special global press is the tense of RDBMS to cover traditional GIS territories. These efforts target to reduce the high cost of a classic GIS solution, designed by other age's philosophy. Until then, information was not accessed by the internet, so GIS software was addressed to minimal user licenses. This reaches nowadays as a very expensive cost of GIS software per user license. This logic has been correctly replaced by the tense of Information to be wide accessed by the internet, increasing the total number of end users. The internet has filled the technological gap, in order to establish total and wide access to spatial data. So, classic LIS solutions that until now were valued by the total number of end users are getting two significant disadvantages: the extremely high cost and the technological inefficiency for very large end users and datasets. Recently, IT has captured this tense and wisely is trying to achieve this relative share of the software market. In result, efforts are being made to reduce the authoring cost and to support spatial data operation. The result is "open source" database applications, metadata support inside the database core, maps generated by application servers and spatial information analysis using SQL queries by the database engine. So, if someone is willing today to develop a LIS solution for a modern Organization, he has to consider, besides the alternative capabilities offered, the modern tense and needs. So, a primary goal in such a solution is to support very large datasets and end users, despite a familiar user interface and operation procedures that classic GIS software offers. This is wrong by both technological and financial perspectives. What has to be considered is the trustworthy operation with very large number of users and datasets, which is the subject of this Ph.D. research. In addition it has to be considered that society does not fund such solutions any more with the huge amounts of money that was common in the last decade. Even through funding programs of the "Information Society", budgets for developing geoinformation systems for government Organizations are much lower than the ones of earlier times. In addition, these solutions focus on the application subject rather than software supply, confirming the above mentioned. For instance, the recent Invitations 146 and 148 by the Information Society for Local Administration Organizations were funding LIS supporting smart solutions, like fleet management. The budgets there were so low, that they practically "photographed" open source solutions, not leaving financial limits for the especially expensive software of the classic GIS market. It is true that the modern tense of IT to conquer geoinformatics is unquestioned. In the near future, what will interest are Geoinformation itself and its distribution to the end users, rather than a familiar user interface and work procedures. And it is certain that this geoinformation will have to operate on-line. The term of on-line is very important, because it demands technological innovations to support it. It is logical that an on-line Database demands a serialization preservation mechanism. In simple words, the Database needs not to have unresolved conflicts. And that is usual in common editing of information by different users, when conflicts are created and especially in long term transactions. Until now, the main issue of serialization preservation in a Database is tried to be solved by transaction management mechanisms. Main efforts that are tried the last years by software vendors have been established such mechanisms. But none of them satisfy the issue of serialization preservation and the on-line operation. And this is a fact because these mechanisms demand human interaction in order to direct the machine in conflict solving (Optimistic- Pessimistic Locking scenarios). A relative article of Han Wammes (Market Development Director - Oracle) in Geoinformatics, September 2006, describes clearly the tense of IT to conquer the GIS market and it concludes: "Yes, we are almost there, at least for the basics, but there are still issues to be solved: what about web services, security and transaction management?" This Ph.D. provides an answer to this question.

Στις μέρες μας, οι Τεχνολογίες Γεωπληροφορικής περιστρέφονται γύρω από εμπορικά λογισμικά & έτοιμες λύσεις, ικανοποιώντας έτσι σε αρκετά μεγάλο βαθμό τις λειτουργικές απαιτήσεις Διαλειτουργικών Οργανισμών που διαχειρίζονται χωρικά δεδομένα. Η σημερινή, όμως, τεχνολογία του Γεωγραφικού Συστήματος Πληροφοριών, Geographic Information System (GIS) επιτρέπει σε μια κολοσσιαία Χωρική Βάση Δεδομένων (Β.Δ.) ενός Οργανισμού - που απαντάται σε πολυχρηστικό περιβάλλον εργασίας - περιορισμένες λειτουργίες στην από κοινού επεξεργασία τόσο της ίδιας της Β.Δ. εν γένει, όσο και κοινών εγγραφών ειδικότερα. Μία τέτοια κοινή επεξεργασία συγχωρείται στις κλασσικές GIS υλοποιήσεις μόνο σε ένα επίπεδο συνεργασίας μεταξύ των χρηστών με τρόπο, ωστόσο, που καθιστά την επεξεργασία της ίδιας εγγραφής «συγκρουσιακή», ώστε να απαιτείται επίλυση από τρίτο χρήστη. Το φαινόμενο αυτό συναντάται συχνότατα σε Οργανισμούς που διέπονται από την Αρχή της Διαλειτουργικότητας, της διανομής δηλαδή της πληροφορίας τους εντός και εκτός Οργανισμού, και οι οποίοι διαχειρίζονται κολοσσιαία sets χωρικών δεδομένων με συναλλαγές μακρού τύπου (long term transactions). Η παρούσα διατριβή μελετά τις Σύγχρονες Χωρικές Β.Δ. (Spatial Databases) και αναλύει την εφαρμογή τους, ως ένα GIS επιδεκτικό, αυτόνομο, βασισμένο στα διεθνή πρότυπα, το οποίο φυσικά διέπεται από την Αρχή της Χωρικής Διαλειτουργικότητας. Παράλληλα, εξετάζει τα φαινόμενα που δημιουργούνται στις πολυχρηστικές αυτές χωρικές Β.Δ. και εμβαθύνει στην δημιουργία ενός αλγόριθμου: του Αλγόριθμου «FoCHAL» (Free of Conflicts Hierarchic - Altruistic Locking algorithm). Δια του αλγόριθμου FoCHAL γίνεται εφικτή η παράλληλη επεξεργασία κοινών εγγραφών από διαφορετικούς χρήστες, διατηρουμένης ακεραίας της σειριακότητας της Β.Δ. με κατάλληλη επίλυση των προβλημάτων και, μάλιστα, πλήρως αυτοματοποιημένα. Η νέα αυτή Τεχνολογία Γεωπληροφορικής εφαρμόζεται στο επιχειρησιακό μοντέλο λειτουργίας της χωρικής Β.Δ. του Εθνικού Κτηματολογίου ως ένα ανοιχτό και διαλειτουργικό GIS. Δηλαδή εκεί ακριβώς όπου αναδεικνύεται σαφέστατα ο αναπτυξιακός χαρακτήρας του Συστήματος του Εθνικού Κτηματολογίου, γεγονός που συνεπάγεται τεράστια λειτουργικά, κοινωνικοπολιτικά & οικονομικά οφέλη. Το σημαντικότερο είναι η διατήρηση της σειριακότητας της χωρικής Β.Δ. και η επίτευξη της «On-Line» λειτουργίας της. Αυτή η λειτουργία επιτρέπει την ανάπτυξη ορθολογισμένων SubLIS, διαδικασία που πηγάζει από τη χωρική διαλειτουργικότητα κι επιτρέπει την επίτευξη της ηλεκτρονικής διακυβέρνησης.