نویسندگان
1 دانشگاه آزاد اسلامی واحد تبریز - گروه مهندسی کامپیوتر
2 دانشگاه تبریز - دانشکده مهندسی برق و کامپیوتر - گروه مهندسی کامپیوتر
3 دانشگاه آزاد اسلامی واحد خسروشاه - گروه مهندسی کامپیوتر
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
Type errors are one of the main problems in computer systems and database systems. Existence of type errors within databases, not only causes accuracy problem for database, but also leads user to re-enter the record into database because the entered record could not be found. It results in redundancy. The existing error detection methods are based on dictionary. It means that they use a large dictionary whose all words are assumed true and if a word is not in the dictionary, it is detected as a type error. Providing a large dictionary with high precision is expensive and time consuming. In addition, such a dictionary belongs to a special field (for example, medical environment) and is not applicable in other fields (such as sociology). In this paper, a method is presented that could detect type errors without requiring a dictionary. The proposed method has been evaluated with some common criteria. The experimental results show 93.5 percent precision for this method. In addition to the high precision of the proposed method, not requiring a dictionary is considered as its unique feature.
کلیدواژهها [English]