In recent years, the promotion and monitoring of physical activity has been the focus of public health work. However, it is challenging to objectively measure the level of physical activity in the population, because it requires the use of expensive equipment to track a large number of people and impose strict data collection protocols. In other words, emerging technologies can provide reliable and effective alternative monitoring tools for self-reported physical activity measurements. Although population levels of physical activity data collected using integrated sensor technology research quantity increased, but want to use this technique in the study of its researchers, there is little technical guidance.
According to graham and heep, “physical activity measurement research achieves greater ease of use, accuracy, and scope by integrating emerging technologies. These emerging technologies are worthy of attention because they can be used to improve the external effectiveness of measures and results through ease of use and transferability; The ability to improve the analytical model significantly; Improve the systematic collection and analysis of public health supervision due to real-time capability; And it needs to study the network infrastructure needed to process big data.
In 2010, the U.S. department of health and human services has released the fifth part of the national institute of health and health report, reflects the scientific and powerful state support for the benefits of regular physical activity on health, it is based on the previous results of healthy people initiative. The report, entitled “healthy people in 2020”, introduced the new decade goals for health and disease prevention. The goal novice is myHealthyPeople, a challenge for technology application developers. The research discussed here reflects the efforts to meet this challenge. The use of the fitness twitter classification model developed for this study enables researchers to collect data in real-time, which is in sharp contrast to the telephone interviews that rely on participants’ recall. The biggest challenge in using technology to track physical activity is to show that many users are inconsistent in using the tracking device.
As part of a “healthy people 2020” includes physical activity, which suggests that americans should at least once a week for 150 minutes of moderate-intensity physical activity, in order to obtain substantial health benefits, and more than 300 minutes a week to get a wider range of health benefits.
In 2008, when healthy people 2020 goals and objectives of the development for the first time, 43.5% of American adults reached 150 minutes a week of moderate-intensity physical activity goals, only 28.4% of 300 minutes a week.
For this project, we open to using Twitter share of healthy people 2020 report to collect data of physical activity and move from five fitness applications (Nike +, DailyMile, MyFitnessPal, Endomondo and RunKeeper) to collect data of physical activity were compared.
This study used in each mobile fitness applications have a standard word phrase, used to use the fitness tweets automatically share physical activity, including physical activity time and/or distance. In addition, some mobile fitness apps are included in standard phrase phrases, and the shortened URL points to the user page of the mobile fitness application.
On this page, you can collect additional information not included in the exercise tweet. Write a data grab script to collect this information. Once the data collected, the data cleansing can eliminate the low total (28 weeks less than 15 minutes of physical activity report) and high total more than 30000 minutes of physical activity (28 weeks report), to a one-time consumer or user errors, and is the result of technical problems related to invalid results (for example, after the completion of the application in exercise fitness keeping open, making digital expansion and make the data skew).