- Public Sector
- Customer Service
Social Monitoring: Using AI to analyze article sentiment
It is often difficult to identify when negative articles or comments about your company are posted online. Enabling you to act and respond quickly, Ciphix offers Social Monitoring. Powered by AI, the Social Monitoring procedure analyzes the sentiment of sources such as blogs, news articles or platforms of certain sectors, generated by a Google Search. The result is summarized in a report containing the source URLs and webpage-sentiment scores.
Use case: Identifying negative comments by hand
The Relationship Management department of an unemployment benefits foundation tries to identify negative articles or comments about the foundation online. Citizens often make requests for additional unemployment benefits from this foundation. However, granting a request follows a specific procedure to determine eligibility. People find the application process rather difficult, and often express their feelings about the process online via news articles, social media platforms or blogs. While there are plenty of positive comments about the foundation and their support, there are also negative articles and comments. Unfortunately, not all buzz on the internet can be positive.
If the foundation cannot identify these negative articles or comments quick enough, it can lead to negative escalations. As a result, it is challenging for the foundation to give the complainants the proper support and consequently improve their own image.
Identifying negative comments in sources such as blogs or news articles is a repetitive and time-consuming task. It is also difficult to act timely enough to resolve conflicts, as the process of identifying negative comment threads is lengthy. For this reason, we implemented an AI solution called Social Monitoring for the foundation. This procedure automatically analyzes the sentiment of sources such as blogs, news articles or platforms of certain sectors, generated by a Google Search.
Our solution: AI as a Social Monitoring tool
Ciphix provided a social monitoring tool that acts as a ‘hawk eye’ on the internet, looking out for relevant content. The tool uses the Google Search Engine to perform certain queries which range between certain periods. Next, our solution scrapes the content of the resulting websites. This content is analyzed, and the tool generates an overall document score between 0 and 1, as well as individual sentiment (positive, neutral or negative) percentages. The results are exported to an Excel-sheet and e-mailed to the relevant stakeholders. This social monitoring by Ciphix ensures that a new hit or mention does not need to be analyzed manually but provides the article sentiment automatically.
Our solution is built to take the following steps:
- Performs queries in Google Search Engine twice a week
- Scrapes content of resulting websites and analyze the content
- Generates overall document score between 0 and 1
- Generates individual sentiment (positive, neutral and negative) percentage
- Sends e-mail with report (including source URLs and sentiment scores) to department
The process of analyzing sentiment of content is 100% automated. The Social Monitoring solution ensures the capturing of insights that otherwise would have gone unidentified.
