Data Mining Technique And Evaluation In Iraqi Named Crime Documents

1Hassan M. Ibrahim, Methaq A. Shyaa, Ali N. Yousif, Alia J. Ouda


Named entity recognition (NER) products attempt to instantly understand and also classify the proper nouns in text that is written. NER devices possess a significant component in a lot of areas of Natural Language Processing (NLP) like as issue answering methods, text summarization and information retrieval. Unlike previous Arabic NER approaches which are created to acquire called entities from fundamental Iraq textual content, our method entails removing named entities from criminal newspapers. Extracting called entities from criminal textual information gives basic information for criminal analysis. This paper offers a principle based strategy to Iraq NER os appropriate to the crime url. Based on morphological information, predefined typical indicator lists and also crime as well as an Arabic named entity annotation corpus from criminal url, a lot of syntactical rules in addition to patterns of Arabic NER are triggered then formalized. Then, these rules and patterns are used to discover as well as classify named entities in Arabic criminal information. The end result suggests that the accuracy of our product is 94 %, which conclusion implies the method functions as well as the performance on the achieved unit is positive.


Named entity recognition, Natural Language Processing, Semantic Inferential Model, Part of speech

Paper Details
IssueIssue 6