This research project aims to investigate the relationship between memory and information retrieval methods within recommender systems. This will be carried out using a mixed-methods process: through a survey, interviews and user testing. This topic is important because the recommender apps are very widely and increasingly used which make them an important topic of study. In addition the cognitive frameworks, the recall and recognition influence user behaviour and decision-making. I hope to find out how the user’s memory load can be reduced as a result of information retrieval and interface design.
Goal: Investigate information retrieval methods and their effects on the design and structures of mobile interfaces through user interaction with recommender systems in order to decrease their memory load and increase their decision-making abilities.
Objectives: Recommender systems exist in a variety of applications. The domains that users are familiar with include some of those that are extremely popular for the use of mobile apps, such as restaurants, e-commerce, travel destinations and financial services. The construction and adaptation of filtering and search techniques in these interfaces lead users to discover information which they might not have found by themselves. The study will identify how users employ these techniques to reduce the stresses associated with the task of finding products to purchase, and measure how frequently they need to recall the associated information from their memory. In conclusion, we will find out how different methods of retrieving information in the recommender apps support decision-making activities.
Requirements: The aim is to compare of a variety of mobile websites and apps, which will be tested according to the user’s perception, memory and reasoning abilities. The smart phone’s usability is crucial. People’s use of, and points of view on, web browsers and search queries will be analysed first via a questionnaire and then in greater depth via interviews. For the interviews, data will be gathered from users using constructed scenarios and tasks. The survey will be sent to max 100 people and 10 of these will be selected from among the people who have responded to the questionnaire.
Limitations: The study is limited to recommender apps for e-commerce and shopping sites because of time restrictions; in addition, as the investigation of information retrieval methods in financial services may include sensitive data, it may be difficult, both in practical and ethical terms, to carry out a project that involves research into the user behaviour associated with money transfer apps. As a result, the study will focus on the information retrieval methods which are applied in recommender systems for shopping and e-commerce sites (Ricci, Rokach and Shapira, 2011).
1. How do different methods of retrieving information in recommender systems differ in terms of recall and recognition?
2. Why do these differences exist, and how are they reflected in the structure and design of a recommender system?
3. Do these different methods, structures and designs influence user decision-making and the load upon their memory?
Systematic usability testing will be undertaken. The grounded theory will guide me in terms of data collection and the many subsequent steps for data analysis. I will generate my version of the grounded theory during my research process. The key steps consist of user research, conceptual design and evaluation (Morse, 2009).
1. User Research and Demographics
Questionnaire and Data Analysis – Quantitative Approach
First the ground theory will be undertaken to understand the patterns of the users’ views and ideas about searching for information, and different mobile website and app designs will be compared. I have chosen to analyse mobile interfaces because nowadays, users increasingly access the information they need by using smart phones; moreover, the use of these media in such an analysis is more convenient to collect a high volume of data, and the evaluation will be quicker in person. To start, I will use a questionnaire to gather a range of data. The questionnaire conducted will display the subjective assessment of users based on the cognitive frameworks. The answers will be compared as result of data collection and scenarios. The time constraint sets up after selecting users according to their previous experiences with recommender systems. The use of mobile phone may create a storage problem because the mobile apps take a large storage space therefore other mobile interfaces (websites) will be considered.
• Information Research: I plan to test the first and second hypotheses. The questionnaire will be used to see how users behave at the moment of interaction with information through digital interfaces, for instance: whether or not they check websites, mobile websites or mobile apps? If so, why do they choose the specific ones? What are their preferred websites or mobile apps? If they use a website, do they prefer a mobile version or they just go to the equivalent mobile app directly? How do users extract data when they deal information overload? What are the main factors which affect their decision-making? As a brainstorming exercise, I will ask users to perform card sorting in order to categorise objects to understand how they organise information in their minds. The goal is to identify user demographic and itinerary details before the task is given.
2. Conceptual Design
Scenarios, Natural observation, Semi-structured interviews – Qualitative Approach
After I gather data through the questionnaire, I will evaluate the mobile interfaces upon the users’ requirements; this is fundamental for a user-centred design. I will identify suitable respondents from the questionnaire and conduct interviews with them in order to analyze the differences and similarities of the designs of mobile interfaces; the questionnaire will then be re-evaluated based on the usability features present in mobile digital interfaces, such as navigation, filters and search results. The interviewees will be observed carefully. The aim is to evaluate the interaction design according to factors such as simplicity versus complexity, consistency versus vagueness, and familiarity versus novelty. The aim of conceptual design is to describe what users can do with a product. The interviews will be taken place in the screens of interviewer’s mobile phone. Meanwhile the mobile apps require a large storage place so mobile websites can be also used during the interviews. The questionnaire and semi-structured interviews produce a variety of raw data, which I will use in conjunction with a comparison of existing interfaces such as e-commerce, book and shopping recommendation apps and event/hotel-booking websites. It is important to make a comparison of the functions which are based on user journeys and interface sitemaps. This qualitative analysis will identify and present the subjective views and design patterns. The goal is to identify task involvement and user behaviours. For example, can we observe that the more the quantity of filters increases, the less the demand on memory occurs?
3. Detailed Design – Approaches to User Behaviour
After the quantitative and qualitative stages are completed, the data collected and analysed within its theoretical context will be used to create a technical design. According to the cognitive approach, the environment and social experiences let users to process abundant information and behave as a response to the stimulus (Schmitt, 2003). For future work and study, the wireframes/mock-ups of mobile websites and apps will be drawn to observe the stimulus caused by the interaction between a real, tangible environment and the user’s social experiences within a visual interface. The design of a new prototype will be drawn according to the simplicity, clarity, novelty, and visual consistency factors. Paper sketches, Balsamiq and JustInMind tools are my current preferred prototyping tools depending on the ease of use and timing.
4. Evaluation – Results of Findings and Critical Discussion
I will conduct evaluation sessions to finish data gathering, and then display the results. Meanwhile, the project will be completed with the design of sampling charts and tables of the numerical values for the quantitative data, and the design of coding to highlight headings and analyse the qualitative data. The coding will be performed with Maxqda tools to achieve a visual demonstration of the data analysis (Maxqda, 2014).