FINAL EXAM Flashcards
(25 cards)
- Collection & categorization of personal data associated w/ a specific user
- Data Structure used to capture certain characteristics about certain user
User Models
Actual representation in a given user model
User Profile
- Process of eliciting stakeholder needs and desires and develop them into an agreed upon set of detailed req.
- make the problem stated clear and complete
- ensure that the solution is correct, reasonable, and effective
Requirements Engineering
- More technical approach to:
1). Defining the target audience
2.) Adapting the system to the user needs - Subdivision of HCI used to create understanding of a user
User Modeling
Create reasons for the development of the software that the user accept, that is:
1. flexible enough
2. Open to changes
3. abide by standards
Feasibility study
- current technologies are evaluated using this
- necessary to achieve the requirements with the given time and budget
Technical Feasibility
- assessment of the range for the software
- software preforms series of the levels to solve the problems
Operational Feasibility
Starting step of requirements analysisq
Elicitation of requirements
Models are used DFD, ERD, and FFDs
Specification of software requirements
requirements that are laid down in the document are validated
Validation of Software requirements
new requirements w. the change in the needs of business
Management of Software requirements
Find out what people are looking for to adopt your system
Relevance in feedback
you need to adopt the system for mobile environment to see if the user behavior changes
Mobile environment user modeling
- one of most traditional ways
- Demographic research statistics (Education, religion, age, gender, and etc.)
- different people groups respond differently to different stimulation
Demographic User Modeling
- remains static and unmodified
- useful for services that do not need customization immediately (e.g. blog)
Static User Modeling
- information is gradually updated (e.g. Recommendation engine)
Dynamic user Modeling
- created from the generalized version of static profiles
- make assumption based on larger chunks united by common characteristics
- opposite of HAA
- useful when you don’t need / have access to personal info
Stereotypical User Modeling
- uses a labeled dataset
- can detect anomalies in the data
Supervised Classification
- Algorithm looks for relationship and values between different elements
Supervised Regression
- trained on the go by looking for patterns and clustering similar information
- Unsupervised Clustering
- Multiple decision trees help to assign different data
Random Forest
- Created to adopt the content the person sees
- Beader - Meinhof Phenomenon
Social Networks and Search Engines
Learning how your current clients uses the product and service
Product Management and Improvement
- to create powerful marketing campaigns
- learn from the data available
Digital Marketing and adtech
