[2022-2학기 DSL X Visual Camp 기업연계 프로젝트] 시선데이터 기반 읽기 능력 평가 서비스 개선

[2022-2학기 DSL X Visual Camp 기업연계 프로젝트] 시선데이터 기반 읽기 능력 평가 서비스 개선

팀원 강민주 김원 유채원 이승주 조보경 최명헌 황진우 전혜령 김예진

Notice

Some codes and data are not available because of the policy of company.

Overview

Motivation

Problem Definition

How to run

python main.py

There are some arguments to be changed.

• is_sample = True : Run SmartVC with specific sample. You can see more details at initialization of DataHandler(utils/data_handler.py)
• sample_id = 0 : In the case of run SmartVC with specific sample, you can choose which sample to run.
• export_table = True : Once we run SmartVC, we can export the result of the whole data in xlsx file. There are 3 files to be exported, which are tables of CorrectedFixation, RawFixation and WordAoi.
• corrected_line_x(y) = True : Run line allocation algorithm with correction(our suggestion). If you allocate False, then line allocation is based on Nearest Neighbor Algorithm. Or, you can run with our suggestions.
• log_all = True: When you run SmartVC, you can easily debug current status.

Details

Data Structure

We wrap raw data with Python object. class Visc has several member variables, from Raw Fixation to WordAoi.\ You cannot see more details because of the policy.

Primary Modules

• data_handler.py

This is for DataHandler of our data. You can see more details regarding data process.

• iVT.py

This is compose of several function to run iVT Filter. We can tell the process of iVT Filter in the function named "run".

Sample