How To Rank Internet TV Content
Secondary factors:
1. Vote (star) (score)
2. Comments (Comment)
3. Visit duration
Primary Effective Factors:
1. Like
2. Dislike (dislike)
3. View
4. Upload_date (Loading Date)
On the tv.fam.ir site, the direct and indirect contact of the audience is as follows:
1. View the content
2. A review (which, of course, does not allow for commenting by registering and logging in the site) (the number of words is also considered in the calculation)
3. The rating (the star) of the content (which for the time being only management can give this rating to the content)
4. Time to view content online (most likely this item is still not trackable and not registered)
As seen, there is no likelihood of likeness (dislike) or dislike.
According to the above, a formula for content ranking in various fields should be prepared, the scientific and exact form of this data-mining model is complex over time and requires specific and localized thematic studies (some of these articles It means that you can not even be modeled from other sites and replace them. According to the feedback from the audience, FAM.ir has to do this at least 3 months later, on The basis of this study is the data mining of the initial formula coefficients and the number and the limits of its parameters, and may interact The audience also extends to other fields (such as the location of the mouse on the page, the number of clicks in different areas of the page, and other logs that are not important in the first phase.
Therefore, the initial formula for this ranking is as follows:
The average of scores for all content as a constant coefficient affects the overall result of the score, so it comes in the form of a coefficient and increases the rank with its rise (direct impact). The mean value of the visit time that shows the power of the content’s impact. Multiplied by the number of hits. (Direct Impact) The average number of words in a comment that shows the challenging content is multiplied by the number of comments. (Direct Impact) Since everything from the date the content is being transmitted has a reverse effect on the rank and leads to a lower ranking. This time interval is loaded up to the current hour on the whole. Since the effect of the loading time and the growth rate Its impact on other things should be higher. This item is not a coefficient of influence and is calculated as power.
These weights can be calculated as the following integers in the initial state:
The baseline weight is for visiting, over time and in response to feedback from the audience, these weights should be modified according to the needs of the respondent according to the data mining algorithms. The corresponding initial weights may also be changed for any content in the first instance, so entry of meaningful content and reporting should also be made at the outset.
