Gold Mine in Users’ Opinion
There are huge UGC (user generated content) online nowadays. Some existing tools can conduct sentimental analysis to these UGC and get a sentimental tendency ranking. On the other hand, there are lots of website analysis tools that can analyze user browsing/purchase behavior and generate user behavior reports for the website. Based on these reports, website analyzer can measure how effective the website is and then optimize to reach a better convention rate. Now we consider to integrate user sentimental analysis result with user behavior analysis report together to get more information about products, website, or users to help web analyzer to do the improvement. As a web analyzer or expert in this field, do you think this would be helpful for you? What requirement you may have on this?
Scenario 1: Current sentimental analysis tool can analyze user’s comments posted in the website and get user attitude towards the product. For example, we can get user comments to a type of camera, shown in the chart below,
After basic sentimental analysis, we can find the most outstanding positive comment is its high resolution, and the most outstanding negative comment is that the lens performs not so well. But this result can’t tell analyzer which user group contributes to these comments. Considering if we combine this with other data obtained from existing web analysis result, like user age, gender, location etc, then we can get more meaningful and instructive result like below:
In this result, we can find customer of age 50 has different attitude from customer of age 20-30. Age over 50 group is more satisfied with resolution and manufacture, but age 20-30 group likes slim size and rich functions. If we can get this kind of result to indicate different attitude to a product from user groups of different age, gender and location, do you think it can help you in website optimization?
Scenario 2: The combination of sentimental analysis with behavior analysis can also be applied in more accurate user grouping. Currently website analysis only focuses on user behavior analysis. We can get more detailed user targeting results by adding user comments/sentimental analysis. For example, we have 2 customers John and Anna who bought same product and left different comments in website. John mentioned he feels the price is high but the performance is good. Anna said she likes the fashion outlook and price is ok. From these different opinions, we can mark them with different tags or divide them to different user groups. Targeting to each groups, different product will be recommended. For example, for John’s group, you can recommend cost-effective products which is more possible to get ordered. Please look at the analysis result example below and see if this would be helpful to you.
Hope you like it! Thank you!!