AI in Starmind
Starmind's patented algorithm is unique and active in many areas of Starmind to make your network more powerful, connect the right people at the right time, and find out about each employee’s unique skills.
Read the introduction below, or jump directly to one of the following topics:
Artificial intelligence in Starmind
The field of artificial intelligence combines insights from computer science, neuroscience, statistics and linguistics to create computer programs that simulate or mimic human intelligence. Starmind uses artificial intelligence to empower employees and organizations to use their full potential by identifying expert knowledge, making undocumented knowledge accessible and developing talent intelligence. Starmind combines state-of-the-art methods from public artificial intelligence research with our own in-house research and development in order to achieve the best results for our use cases.
The use of artificial intelligence in Starmind can be divided into two main areas:
- Starmind automatically extracts the most relevant topics from different types of documents and understands how these topics fit into a skill ontology by using techniques from the field of natural language processing.
- Starmind builds skills profiles by learning from different kinds of interactions of users with topics, and continuously updating this information in real time as new relevant data comes in.
Natural Language Processing
Starmind uses a variety of methods from the field of Natural Language Processing to identify the most important topics in questions, answers and different types of documents. In the Starmind application, these topics are represented by tags: single words or short phrases that describe a topic of expertise. These tags can come from general knowledge that is preloaded into Starmind. Additionally, Starmind can also recognize organization-specific tags, such as project names.
While tags are visible to the end-user, they also have limitations, because some details are lost when reducing the full content of a question or other document to a limited number of tags. For this reason, neural network models are also used in the background to maintain a more fine-grained representation of how different topics and documents relate to each other. These models are able to keep much more detailed information about the original text, while at the same time they are much easier to use in algorithms compared to the original plain text.
Learning about skills and expertise
Expertise is Starmind’s internal representation of the skills and knowledge of each user. A user’s expertise on any topic can grow or decrease depending on how the user interacts with this topic on the Starmind Q&A application or on other connected applications (see " Expertise and contributing factors”). The more different sources Starmind learns from, the more accurate and complete Starmind’s expertise profiles become (see “Connectors”).
Starmind’s expert search algorithm uses expertise to determine who is an expert on a given topic. Expertise is also used to select the most relevant topics to be shown on the skill map of each user.