What I think about the recommend system

After the class we learn last week, I have a new idea about how we make decisions with the help of recommend system. We usually rely on some suggestions and recommendations when we make decisions, and we face many information that we can not breath well. With the information overload, how can we make a good decision and what we can we make to filter information, and identify items which are relevant to us.

There are three ways that we can use, respectively they are recommender systems, recommendation systems and recommenders. First of all, I want to talk about the recommendation systems. There is an extensive class of Web applications that involve predicting user responses to options. Such a facility is called a recommendation system. The goal of a Recommender System is to generate meaningful recommendations to a collection of users for items or products that might interest them. Suggestions for books on Amazon, or movies on Netflix, are real world examples of the operation of industrystrength recommender systems. The design of such recommendation engines depends on the domain and the particular characteristics of the data available.

There are also three types of recommender systems. They are contentbased recommendation, collaborative filtering recommendation and hybrid approaches. 

Contentbased recommending: These approaches recommend items that are similar in content to items the user has liked in the past, or matched to attributes of the user.

Collaborative Filtering (CF): In CF systems a user is recommended items based on the past ratings of all users collectively.

Hybrid approaches: Thesemethods combine both collaborative and contentbased approaches.

As we continue to integrate more information into recommender systems, and as we expect them to adapt to greater contextual challenges, the problems of algorithmic and system scalability only grow. A key challenge to the field of recommender systems as a whole is the integration of contentbased approaches, collaborative approaches and contextual approaches into comprehensive, practicalrecommender systems. We are already seeing indications of strong interest in this direction. One longrunning challenge is to recommend the next item in a sequence of items to be explored. Such a recommender requires not only ratings of the individual lessons, but also information on sequence and dependency. Similar challenges exist with other forms of context and with recommenders that hope to tap the wide range of implicit behavior that can today be recorded through web browsing behavior.

The important of graphs in social network analysis.

In lecture six, we learnt that it is important to do network analysis with the help of graph. Because social network is a common component in social media, and computationally, social network can be viewed as a graph that describe the relation among a group of people. So if we can do a good graph about weibo or tweeting, we can make a good understanding of the social media. Thus, graph is a very good instrument in social media.

As Rosanna told us that a graph contain a collection of vertices and edgs, and there are directed graph and undirected graph, we can know some characteristics about graph, and then we use it to do social network analysis. As Roasnna has told clearly about the characteristics of graph, so I will not repeat it at all. 

First of all, I want to talk about the relationship between graph and network. Graph and networks are all around us, including technological networks ( the internet, power grids, telephone networks, transportation networks,...), social networks (social graphs, affiliation networks,...), information networks (World Wide Web, citation graphs, patent networks,...),biological networks (biochemical networks, neural networks, food webs,...), and many more. Graphs provide a structural model that makes it possible to analyze and understand how many separate systems act together. The Wolfram Language provides state of art functionality for modeling, analyzing, synthesizing, and visualizing graphs and networks.Whether those graphs are small and diagrammatic or large and complex, the Wolfram Language provides numerous highlevel functions for creating or computing with graphs. Graphs arefirstclass citizens in the Wolfram Language; they can be used as input and output and they are deeply integrated into the rest of the Wolfram Language.

e

Then, I want to talk about the social network in the real word and the use of graph. A network is a set of items (nodes or vertices ) connected by edges or links, and the systemsis taking the form of network abound in the word. Thus, we can usually learn a lot about people from studying their social network by using graph. So the social network can be represent by a lot of new graphs and they are generally noereflexive. Many social network are all graph. The web itself can be viewed as a very large graph, nodes are individual sites or pages and edges are the links between pages. The following picture is the basis of Google’s page rank algorithm. The ‘importance’ of a site is determined by the number of sites that link to it weighted by the importance of those sites.

From the PPT we read and the papers we reference, we found that graph can help us do a lot of meaning things and know clearly about the social connection among people. 

What's the meaning of Sentiment Analysis and Opinion Mining

In the last class, the teacher said something about Sentiment Analysis and Opinion Mining. In order to know clearly about it, I read some books and look up some information about it.

We always chat with each other to exchange opinions, which means our beliefs and knowledge are expressed externally through our opinions, they are central to almost all human activities and are key influencers of our behaviors. In order to be clearly about people’s opinion, we need to do sentiment Analysis, which means we need to do computational study of opinions, sentiments, appraisal, and emotions expressed in text. It is also known as Opinion Mining. We always express our opinion on blogs, discussions, microblogs and social network.

Sentiment analysis could help us do a lot of surveys. For this reason, when we need to make a decision we often seek out the opinions of others. This is not only true for individuals but also true for organizations.

From this picture, we can know what kinds of people are willing to purchase what kinds of things.

In order to make Sentiment Analysis, we have to do holder detection, target recognition and sentiment classification. The basic task of opinion mining is polarity classification. Polarity classification occurs when a piece of text stating an opinion on a single issue is classified as one of two opposing sentiments. Reviews such as “thumbs up” versus “thumbs down,” or “like” versus “dislike” are examples of polarity classification. Polarity classifications also identify pro and con expressions in online reviews and help make the product evaluations more credible. But we always meet some problems, if people can not explicitly express emotional states, and therefore the tools can’t capture a reviewer’s implicitly expressed opinion or sentiment.

Finally, we found some evolution of opinion mining. Currently, opinion mining and sentiment analysis rely on vector extraction to represent the most salient and important text features. We can use this vector to classify the most relevant features. Two commonly used features are term frequency and presence.

The Web has changed from “read- only” to “read-write.” This evolution created enthusiastic users interacting and sharing through social networks, online communities, blogs, wikis, and other collaborative me- dia. Collective knowledge has spread throughout the Web, particularly in areas related to everyday life, such as commerce, tourism, education, and health. Despite significant progress, however, opinion mining and sentiment analysis are still finding their own voice as new interdisciplinary fields. Recent approaches aim to better grasp the conceptual rules that govern sentiment, as well as the clues that can convert these concepts from realization to verbalization in the human mind. Future opinion-mining systems need broader and deeper common and commonsense knowledge bases. More complete knowledge must be combined with reasoning methods that are more deeply inspired by human thought and psychology. This will lead to a better understanding of natural language opinions and will more efficiently bridge the gap between (unstructured) multimodal information and machine-processable data.

Blending scientific theories of emotion with the practical engineering goals of analyzing sentiments in natural- language text will lead to more bio- inspired approaches to the design of intelligent opinion-mining systems capable of handling semantic knowledge, making analogies, learning new affective knowledge, and detecting, perceiving, and “feeling” emotions.

What I think about social media

My name is Hanxiao Fang, and my major  is Information Engineering in Chinese University of HongKong, I recently take part in a course named social media, from the class, I have learned something which I want to share.

Firstly, I want to ask you,what is social media. The best way to define social media is to break it down. Media is an instrument on communication, like a newspaper or a radio, so social media would be a social instrument of communication.But social network is a social structure consisting of individuals or groups who are connected to each other. I think social network belongs to social media. Social media is not only giving you informations, but also let you make commends as well. Think of regular media as a one-way street where you can read a newspaper or listen to a report on television, but you have very limited ability to give your thoughts on the matter. Social media, on the other hand, is a two-way street that gives you the ability to communication too.The latest happenings in social media, plus tips on using Twitter, Facebook, YouTube, Foursquare and other social tools on the web.

Next, I would like to talk about social media analysis.From social media analysis, we can study human social behavior and create better systems to support users and their social activities. We can analysis some character from a large amount of data collection in social media services. From the data we have collect ed, we can through content analysis, network analysis and banking to make some meaningful survey. Rosanna give us some cases about the discoveries we learnt from social media analysis.

The following classes let us know content analysis. From the paper I read on the Internet, I found that we can we can do a lot of meaningful findings which can help people and make out life convenient. From the article I read about "10 things we learned about social media users' charitable habits", we can found the answer about "do social media users really absorb the message?" or "do people use activism on a social platform instead of actually donating money to a charitable cause?" .

There are some founding I want to share with you.

From the picture, we can find that the precise nature of the relationship between social good and platforms like Facebook and Twitter can often be unclear. 

There are many ways we can use to make analysis, like NLP and document comparison which I want to learn more clearly and talk about it later.