Bigram feature extraction and conditional random fields model to improve text classification clinical trial document

A new classification model from the cancer clinical trial document text is proposed to compete with other methods in terms of precision, recall, and f-1 score. Classification of clinical trials text has been developed through several approaches such as statistical approaches [8]; eligibility screening approach [9]; machine learning approach [10]; deep neural network [11, 12] clustering [13], convolutional neural network [14] and approaches to fine grained document clustering [12], It appears that the use of deep learning methods in the modeling of clinical trial classification is still limited. [...]the classification of clinical trial texts is still being developed. [...]it still needs to be developed with other additional methods and features. [...]to answer of this problem, we built a new model using conditional random fields (CRF) and bigram feature as our methodology to improving computational value. 2. Based on this data, the mapping was carried out into the components of the patient's complaints (main complaints, onset, other complaints, information, frequency of attacks, nature of attacks, duration, location, course of disease, previous treatment history, and the consequences of disturbances that arose).