---

Classification and Cataloguing are always thought to be intellectual processes and as the computers are not intelligent, it is presumed that these tasks are human prerogative, if not a curse. The present work could lead to successfully building an automatic bibliographic data entry system, it will be of immense aid in the library automation activity in India and elsewhere, as building bibliographic databases is a very tedious job. A heuristic based Expert System is developed to identify bibliographic data elements from both the title and back of the title pages of books. Firstly the system scan title page, back of the title page and the content of the document. An expert system applying heuristics would identify bibliographic elements like title, author, publisher, etc through lexicographic and linguistic logic. Those elements again counter checked from verso through CIP data and ISBN. The system is tested using representative documents of various kind and from different subject fields. It is found that the system have been successful in more than 85% documents.

---