Automatic Information Extraction Using Hidden Markov Model

The number and size of documents in electronic formats is rising very rapidly and due to the flexibility and ease of use, huge amount of information is stored in textual formats in unstructured way than structured materials, where important information or interesting knowledge is buried in. Information extraction (IE), which is one of text mining approaches, is found efficient in discovering particular entities or relations between the entities in free texts. This book designs, develops and evaluates an Information Extraction system for Amharic language texts, in which it extracts valuable information from the texts and put them on the predefined slots or attributes. Among a number of techniques and methods available for IE, Hidden Markov Model (HMM), a sequential labeling based method, was selected as a methodology. In this book emphasis has been given to the available algorithms in Hidden Markov Model to build an Information Extraction system and finally simple counting of correct annotations and standard Information Retrieval (IR) performance evaluation techniques was used to evaluate the system.

Autorentext

MSc graduate in Information Technology from Graduate School of Telecommunications and IT, and BSc graduate in Computer Science from Haramaya University with Great Distinction and currently working as a software developer at EffectiveUI, a US based Software Company.