Among those topics, I would like to study social recommendation. Social media has changed the way to do traditional marketing and even online marketing. Users nowadays are able to publish and express their views and publish information online with Web 2.0. Oftentimes, this information requires a human perspective to process, for example indicate preference, quote alternative sources, categorize, add supporting materials and assessment of quality. This facilitates the potential users to consider, consume, and apply the information about a product and service. It makes social network users easily to refer to particular ratings and items and to redistribute these opinions to others.
Why do people choose your product? How do potential customers locate your goods? How to convert browsers into buyers?
Reputation! Customers trust opinions and buyers who share their interest. Currently, customers do a lot of research online before making a big buying decision. The comments and reviews of previous customers play a more and more powerful role for retailers and e-commerce sites. As Internet is a chaotic information universe, social reviews and recommendation become more useful for consumers and more profitable for e-retailers.
However, the proliferation of user-generated content and the associated metadata on social multimedia sites introduces new challenges in search, for example unstructured data, vulnerability to spam and noise, short lifespan, much of the content offers little value to the general public. What is worse, some malicious users deliberately post misleading information on social multimedia sites. According to the recent news, Apple app store China branch is now heavily suffering the harassment of so-called Internet mercenaries. Those Internet mercenaries post fake comments on certain apps and push them to the Top popular ones. Now Apple users in China can not trust the recommended apps by Apple store, because those apps very likely cheat by employing the Internet mercenaries.
That is the dilemma of social recommendation. How to utilize its merits and prevent from malicious recommendation is such a big challenge.
That is the dilemma of social recommendation. How to utilize its merits and prevent from malicious recommendation is such a big challenge.
That's true. It is not uncommon for merchants on E-commerce website to execute some transactions artificially to make some favorable comments and rating on their own products in order to attract more consumers. This behavior do have a misleading effect on potential consumers. Such honesty problem is the projection of that in the real life.
回复删除Yes. Chinese consumers are suffering this problem heavily on Internet shopping. Maybe real-name policy can solve it to some extent. But it kind of violates the Internet spirit. I hope effective techniques can someday prevent the misleading recommendation. Then social networks will play a key role in marketing.
删除I also agree with you.It is not a easy way to recommend online..To be honest,I buy nothing else,except books.(Even pirate, it is readable，^-^ )..I think we need some strict regulations in this area, then we will use them,trust them and help ourselves in the future.
回复删除Luckily, online book selling in China is relatively better in terms of the misleading recommendation. I hope we can solve this problem by technical methods in the future.
删除Yes, it is really a fact that our minds are affected by a large number of recommendations made by the website or other people. Just as you have studied, I will read the recommendations carefully when I need something to buy online. And sometimes I found some recommendations which was deleted. Maybe that is a way for the websites to protect their own reputation.
回复删除It is controversial that e-retailers deliberately delete those unfavorable comments. But it is is definitely immoral for Hiring so-called Internet mercenaries to post misleading or fake recommendations. Sadly, that is quite common among Chinese e-commerce platform Taobao. On the one hand, we customers have to keep a vigilant eye on every comments. On the other hand, e-commerce platform operator must try to avoid those happening via technical ways.
删除user-generated content bring a huge-challenge, Ambiguity, to social network. It's hard to determine and analyze ambiguous content
回复删除From technical point of view, the user-generated unstructured data makes it greatly challenging to do analysis and data mining. For human beings, things are totally different. That's why researchers consider social media recommendation (e.g. landscape ranking) to fairly promising. However, users' involvement brings the trust issue. How to prevent the misleading contents generated by malicious users becomes another challenge.
删除Weibo marketing, once the "web celebrity " forwards an advertisement on weibo, they could get good pay from retailers, since that one item could be sold very good after recommended by some so-called web celebrities. So we should take it rationally to the recommendations from web celebrities, do not follow them blindly.
回复删除Yes, that is a typical issue on social recommendation. Customers tend to buy what celebrities' use, let alone their deliberate recommendation. It exists the same issue in TV ads and product placement marketing in movies.
删除Exactly. Nowadays more and more professional malicious users are making "illegal" impact on some certain products for profit. If such impact cannot be controlled, the result of recommendation is no more reliable. Another good news is, the recommendation based on the purchase of your friends, or at least those who have same interest as you, are more convinced, because you know them so it make more sense.
回复删除