This week we learned about search engines and algorithms used in online media. Search engines, such as Google and Bing, are software systems designed to search for information on the World Wide Web. Results are displayed on SERPs, or Search Engine Results Pages. Their purpose is to provide the best results based upon specified search queries and keywords.
The three steps a search engine takes to provide users with what they require are:
Step One: Crawling - collecting data from the Internet and websites on the World Wide Web. about 0.5 billion websites are used for crawling in order to collect data for search engine results. A site must be public and linked in order to be a part of the crawling process.
Step Two: Indexing - sorting and categorizing collected data and storing it in a database. The collected data by the crawler is passed on to the indexer. It analyzes and sorts the data into various categories. Once a user enters their search query or keywords into the engine, it tries to match it with already indexed data sets.
Step Three: Processing Search Keyword - providing the best result from the database. This is where the algorithms come in. It filters by search engine system. Google, Bing, Yahoo, etc all use different algorithms. Search engine websites update their algorithms quite often in order to make searches more relevant to what the user wants. For example, Google updates their algorithm 100 or more times per year.
We also learned more about these algorithms. An algorithm is a series of conditions. If something happens, you must respond with this action. It requires a lot of coding to make an algorithm, including the use of "if... then" statements. With multiple conditions, you get better algorithmic results. Items like Twitter's "trending topics," online recommended stories, top search results, and product recommendations are all brought to you by algorithms.
In the article "Can an Algorithm be Wrong?" we learn about Twitter's algorithms used in their "Top Trends" and how it can be controversial due to the fact that it was not displaying topics that were in fact popular and trending. After the Occupy Wall Street event, people were tweeting #occupywallstreet all over Twitter for days on end. However, it failed to show up on Twitter's "Top Trends" section, which, if their algorithm was correct, should have been there based on the number of tweets sent out with that "#" within the allotted time period. This proves that in the end, the one behind the computer screen at Twitter's headquarters gets the final say in what shows up on that trend list. We also learned about another example of this in class, where one student said that "Justin Bieber" had to be blocked from the trend list due to his continual popularity. It really makes you think about whether or not these algorithms are the best thing for you.
My questions for this week are can all algorithms be trusted? Should we believe the algorithms that are determined by your own decisions, or by other users? On search engines, they try and predict what you would like to search for based on previous searches. But on websites like Amazon or Youtube, you are recommended to buy a product or watch a view based on how much other users liked the product or video. Which of these algorithms are better? Is it unfair of companies such as Twitter to control their algorithms and adjust them accordingly without informing their users?
No comments:
Post a Comment