Social Business Intelligence for Libraries: Searching Unstructured Data, Analyzing Sentiment, and Moderating Commentary Are Crucial to Helping Organizations Get the Most out of the Growing Volume of Social Data
Guerrero, Fernando, Gonzales, Paco, Information Outlook
Our company, SolidQ, has developed a series of solutions to enable our customers to get the most out of social data. The main purpose of this article is to introduce different techniques and solutions we have developed for social data analysis. Some of the techniques may be helpful to libraries, while others may just add some case study information. Social data is a fast-moving scenario and therefore demands a continuous research solution.
The volume of social data available to individuals and organizations is huge. At any given time, a continuous flow of data is evaluating, commenting about, or rating any given service, product or piece of information. Facebook walls, blogs, message boards, tweets, and video Websites are important communication channels for these data.
There was a time when people's opinions were only reflected by journalists through television, newspapers, radio, and other traditional media. Nowadays, no thought or fact escapes social judgment. From government decisions to teenagers' fashion trends, nearly all ideas, decision, laws, products, and events generate a social discussion.
There are different ways to benefit from social data. For example, organizations can analyze social data to understand what their customers think about them and improve their products and services. Every single move organizations make will be exposed to a complete review, both in internal and external social networks, so they need to continuously analyze social data to be prepared to respond appropriately.
Sentiment analysis enables organizations to analyze and understand people's feelings and sentiments about a product, service or event. It can be applied to tweets, e-mails, forums, Facebook walls, blogs, comments, and even social votes.
Sentiment analysis extracts information from data related to the item, event, or activity. Libraries can use sentiment analysis to understand what users are saying about books or articles. They can analyze events or services to understand customer needs and evaluate customer satisfaction.
For example, many libraries are interacting more frequently with faculty members in the educational process. One form of interaction is at online campuses, where students can perform activities and learn from online library resources. Social media play an interesting role in this process, and sentiment analysis can help libraries analyze users' interactions, tones and feelings.
The key point here is not to evaluate a book or article--it is to evaluate and improve the process of interacting with books and documents. Users can comment on exercises, activities or supporting materials, thereby contributing to the flow of social data as part of their learning process. Using sentiment analysis, the library can detect problems and find ways to improve in the future.
What sentiment analysis provides is a way to extract sentiments from comments. It analyzes the comments and provides a summary with aggregations (e.g., "The movie was fabulous" = 80 percent). This example does not reveal that 80 percent of the commenters wrote, "The movie was fabulous"; what it reveals is that 80 percent of the commenters liked the movie, though they may have used different words, like film instead of movie or amazing instead of fabulous. Sentiment analysis explores social comments about your organization to help you identify trending topics, opinions or reviews.
Prior to the availability of sentiment analysis techniques, customer needs and satisfaction levels were analyzed using suggestion or feedback boxes that were processed by humans, with a high cost and added ambiguity. Nowadays, the human process would be prohibitively expensive due to the volume of comments to be analyzed. Automated sentiment analysis, on the other hand, performs a continuous review of what is going on in your internal social networks as well as in external networks. …