a. The records in logical file are brought into some sequence as determined by key in the records. b. A computer is capable of sorting record into a “nested” sequence. c. Sorting is done by a “sorting generator”. This is part of the computer’s software and comprises several sophisticated sorting techniques that are called into use according to the file and the sort requirements. d. The need of sorting has dimished in line with the demise of magnetic tape as backing storage. 2. Merging - Merging implies that two or more files in the same sequence are combined into one file. a. File merging Two or more separate files of similar seconds and in the same sequence are marged together so as to form one file. b. Record merging The records from two or more “input” files, usually in the same sequence, are combined one record in the output file. 3. Matching a. Two or more input files (generally in the same sequence) are compared records against record in order to ensure that there is a complete set of records for each key. b. Masmatched records are highlighted for subsequent action 4. Summanizing a. Records with the same key in one file are accumulated together to form one record in the output file. b. Summanizing usually applies to a file presorted into a certain sequence and the resultant file is in the same sequence. c. Records to be summarized are generally of a similar type. 5. Searching a. Searching is looking for records with certain keys or holding certain data and in some way making a note of these. b. An instance is a search for and count of all records with a debt balance of above a certain amount. 6. Information retrieval a. Information retrieval is the process that involves the bringing together of data from several files. b. Data may also be extracted from several files and combined before being presented as information. Chapter 8: DIRECT ACCESS FILE ORGANIZATION AND STRUCTURES. Question 1. Storage and Access Modes? There are 3 principal modes for storing and accessing accords on a disk or drum: 1. Serial mode: - The record are stored contigously regardless of their keys - The sole way of accessing serial seconds is to search through the complete file starting with the first record. Click to buy NOW! P D F - X C h a n g e V i e w e r w w w . d o c u - t r a c k . c o m Click to buy NOW! P D F - X C h a n g e V i e w e r w w w . d o c u - t r a c k . c o m - It is sometimes possible to partition a serial files thus reducing the search time by starting the search at the beginning of a known partition. - A serial file is normally of a temporary nature awaiting sorting into a useful sequence. 2. Sequential mode: - direct access sequential mode normally involves accessing sequential a file that is stored sequentially. - sequential mode is often associated with a master file held in a certain sequence and updated by a transaction file sorted into the same sequence. 3. Indexed_sequential/ selective_sequential mode - Indexed_sequential is a mode of storage where by records are held sequentially and accessed selectively. - Groups of unrequired records are skipped past. - Indexed sequential files may also be accessed haphazandly. 4. Random modes: - Each record is stored in a location determind from the second’s key by means of an add generation algorithm. - The only erricient way to find a record is to use the algorithm - Random mode is applicable to master files Ads of random modes No index is required thus saving storage space It is a fast access method because little or no searching is involved Transaction do not need storing, thus saving time New records are easily insertly into the random file provided they are not excessive in number Dis The main problem with the random mode is in achieving a uniform spread of records over the storage are allocated to the file Question 2. Direct Access Addressing? - The key of record is used to identify by record - The key of record also is used to decide its storage location(or address) 1. Self addressing: - Self addressing is a straight forwards method because a record’s address is equal to its key’s value - The file is inevitably stored in key sequence Ads of self addressing It leads directly to the wanted record No indexing or searching is required The key itself need not necessarily be held within the stored record- although it generally is Dis The storage space per second has to be the same When records one missing, storage locations related to its must be left empty 2. Self addressing with key conversion Click to buy NOW! P D F - X C h a n g e V i e w e r w w w . d o c u - t r a c k . c o m Click to buy NOW! P D F - X C h a n g e V i e w e r w w w . d o c u - t r a c k . c o m - This method a basically similar to self addressing except that the key required a little processing to turn it into the record’s address - This leads to either a pricise address 3. Matrix addressing - In somes case, it is necessary to find the add of a record held within a multi dimensional matrix of record it’s called matrix addressing. Question 3. Direct Access Searching? - Where as addressing determines the location of a record by using algorithmic methods, searching finds the record by scanning groups of records, and index, or both. - ]The simplest method is to examine every record a file until the required record is found a shortcut is generally desiable. 1. Indexed sequential searching - A cylinder index is created to hold the highest cylinder’s key - Associated with each cylinder is a block index holding the highest key in each block within that cylinder - When searching for a record’s key in the index The cylinder index is examined key_by_key until one is found that is larger than or equal to the wanted key this directs the search to the appropriate block index The block index a similarly examined and the search The block is searched record by record until the wanted record is found 2. Binary searching( binary chopping ) - The key in the index to be binary search must be in sequence and form a complete set - The search starts at the midpoint of the index and then moves half way to the left or right(down or up) depending upon whether are wanted key is less than or greater than the midpoint key - In pracice, the index is unlikely to as convenient as this example because it is not always possible to exactly halve each sucessive move(complete exact holvingis possible only when the total number of keys in the index is 2 0 -1) - The average number of examinations comparisons is (log 2 k) -1 ( k is the number of keys in the index) 3. Block searching - A block is a subdivision of an index. A block is devised to contain, roughly the square root of the number of keys in the whole index - The search is first through the block index to find the appropriate block and then through this to find the wanted key - The average number of examinations is square – root – k (k is the total number of keys) 4. Balanced binary tree searching - A binary tree is a relationship of keys such that the examination of any key leads to one of two other keys - The binary tree is actually in the form of an index containing all the keys together with a directory showing the braches stemming left and right from each key Click to buy NOW! P D F - X C h a n g e V i e w e r w w w . d o c u - t r a c k . c o m Click to buy NOW! P D F - X C h a n g e V i e w e r w w w . d o c u - t r a c k . c o m - Binary tree searching is suitable for an unsequenced file - The search is similar to binary searching in that each key examination holves the rinaining keys, on average Chapter 11: INTRODUCTION TO ARTIFICAL INTELLIGENCE. Question 1. AI? Atificial Intelligence It has three braches 1. Expert systems (or knowledge- base system) - ESs are programs that contain the knowledge of human expert, encoded so a computer can understand it with encated- knowledge seasoning machinism, ES can tackle problem that are beyond the seach of conventionally programmed computers. 2. Natural language systems (everyday native language) - Natural language systems are programs that understand the native language of the user, such as E - The most popular natural language systems are those that act as interfaces to data bases 3. Simple perception systems (for vision, speed and touch) - They can interpret visual scenes and decide if object meet inspection standards and quality control criteria, or move a robot to the proper location ot grasp a part for manufacturing Question 2. Who does the updates? - Updating the knowledge bases is very diffirent when with updating databases because of the difference in the type of information and in the cause and effect relationship contained in knowledge bases - A knowledge in the area, when databases may be modified by a normal users Chapter 12: EXPERT SYSTEMS. Question 1. What is an ES( Expert system )? An ES is a knowledge-intersive program that solves a problem that normally requires human expertise Characteristics of ESs - They solve problems as well as or better than human experts - They use knowledge in the form of rules or frames - They can consider multiple hypotheses simultaneouly Types of ES - An assistant Is the leasts expert or lowest level ESs It helps a decision maker by doing routine analysis and porting out those portion of the work where human expertise is required - A colleage The new discusses the problem until a joint decission is reached When system is going wrong, the user adds more information to get it back on track - True ES Click to buy NOW! P D F - X C h a n g e V i e w e r w w w . d o c u - t r a c k . c o m Click to buy NOW! P D F - X C h a n g e V i e w e r w w w . d o c u - t r a c k . c o m Is a system that advises the user without question There are no practical areas today in which decission Question 2. A ES Life Cycle (ESLC)? - An accepted SDLC for expert systems has yet to be developed There are 6 phases life cycle in an ES 1. Phase1 – Selection of an Appropriate Problem - Phase 1 involves finding an appropriate problem for an ES, indentifying an expert to contribute the expertise - Establishing a preliminary approach - Analysing the cost and benefitsPreparing a development plan 2. Phase 2 – Development of a prototype system - A prototype sys is a small version of an ES designed to test assumptions about how to encode the facts, the relationships and the knowledge of experts - The prototype permits the knowledge engineer to gain the expert’s commitment and to develop a deeper understanding of the field of expertise - Other subtasks in this phase: Learning about the domain and the task Specifying performance criteria Selecting an ES building tool Developing an implementation plan Developing a detailed design for a complete system 3. Phase 3 – Development of a Complete System - The main work in this phase is the addition of a very large number of rules - The knowledge base has to be expanded to full knowledge base appropriate to the real world and the user interface has to be developed 4. Phase 4 – Evaluation of the system - This phase involves testing the system against the performance establised in earlier stages 5. Phase 5 – Intergration of the system - The ES has to be intergrated into the data flow and work patterns of the organization - In this stage, the expert system has to be interfaced with other databases, instruments and hardware. 6. Phase 6 – Maintenance of the system - The maintenance of the ES involves is updating, charging in the system when operating. When operating, more problems occur in the system, so it is necessary to continue take care the system by expert in a fix period of time - So expert system, are so complex that in a few year the maintenance costs will equal the development costs. BOOK II: Computer systems architecture. Chapter 1 – 2: NUMBER BASES. Question 1. Common number bases used in computer hardware operation? Decimal(denary) system: Click to buy NOW! P D F - X C h a n g e V i e w e r w w w . d o c u - t r a c k . c o m Click to buy NOW! P D F - X C h a n g e V i e w e r w w w . d o c u - t r a c k . c o m . generally of a similar type. 5. Searching a. Searching is looking for records with certain keys or holding certain data and in some way making a note of these. b. An instance is a search for and. 1. Indexed sequential searching - A cylinder index is created to hold the highest cylinder’s key - Associated with each cylinder is a block index holding the highest key in each block within. V i e w e r w w w . d o c u - t r a c k . c o m - Binary tree searching is suitable for an unsequenced file - The search is similar to binary searching in that each key examination holves the rinaining keys, on average Chapter 11: INTRODUCTION