Keywords

1 Online Bipolar Decision Making: The Brain Hesitation Gap

This paper follows our stream of research on Serendipity Engineering. Although research exists within the Global Systems Science field and social networks in particular, serendipity identification and engineering is rather a new field. Instead of being lost into massive amounts of data, our aim is to find meaning into both and small and mass data so to discover hidden connections appearing as unexpected and pleasant serendipity surprising events. However, chance favours the prepared. Based on underpinned connections and the ways a user acts upon decision making, our proposition is related to the bipolar decision making process. As such, discrepancies in the research are still to be solved and the results are not encouraging as such research now fails in predictions.

The research aims and objectives here refer to the massive global impact of different predicted serendipity events engineered so the users get involved into a predicted decision making process in social media such as Twitter, Facebook, Instagram etc. Furthermore, manufacturing serendipity involves research on initially identifying positive and negative actions and hidden connections between clicks in both big and small data. As such, there can be possibilities to strategically engage users into predicted processes, aid or even interfere into the decision making process. Our research main aim is the identification, collection and analysis of both big and small data in order to shed light in serendipity connections in chrono-spatial intelligence and even engineer it, called Chrono-Spatial Intelligence for Serendipity Engineering (CSISE). A second aim is to identify tools, methodologies and evaluation techniques to facilitate such deep understanding via Chrono-Spatial Intelligence Learning Analytics. Here, we propose a serendipity engineering model for learning insights anchored in big and small data methods, tools and learning analytics.

Choice is the act of hesitation that we make before making a decision. In this mind gap and when online, we decide click or not to click. Every user is battling with mass amount of information on social media; however, such scanning instead of reading, only does attract the attention to a few postings. Furthermore, the user decides on clicking on even fewer. Such initial hesitation, second thought and then, immediate action before changing his mind involves identifying serendipity and taking the risk, accept, and engage into the serendipity flow or deny it.

Stanford researchers [1] observe the moment when a mind is changed. A new algorithm enables a moment-by-moment analysis of brain activity each time a laboratory monkey reaches this way or that during an experiment. It’s like reading the monkey’s mind. This prediction is possible by calculating the average results from different repeated trials. However, average calculation misses the ways the brain functions requiring moment-by-moment research o small data. The authors agree with the San Francisco neuroscientist Benjamin Libet, who conducted an experiment in the 80 s to assess the nature of free will via an electroencephalogram (EEG). Libet’s experiments showed that distinctive brain activity began, on average, several seconds before subjects became aware that they planned to move. Libet concluded that the desire to move arose unconsciously; ‘free will’ could only come in the form of a conscious veto: what he called ‘free won’t.’

A user decides to click or not to click or even changing his mind on the way. If a series of events are engineered towards the happiness and euphoria produced in serendipity, it is rather possible to imply the demise of the free will. However, the change of mind may also imply a change into a new string of possible events and items to click. As the Internet is a connecting ecosystem, the users decide on the social networks to use, creating the background of their own communication space. As such, interoperability between social networks allows the researchers to study the connections and furthermore, to orchestrate actions so the users create the social media buzz on demand.

2 Serendipity Through Macro and Micro Data Perspectives and Analysis

There is no wide research to shed light and understand the magical serendipity experience, which in social networks used for learning is a user-learner experience. As such, serendipity implications for user/learner-centred design processes are very limited or even missing.

In this research proposal, our main research aim is the identification, collection and analysis of both big and small data in order to shed light in serendipity connections in chrono-spatial intelligence and even engineer it. A second aim is to identify tools, methodologies and evaluation techniques to facilitate such deep understanding via Chrono-Spatial Intelligence Analytics [2] in order to manipulate bipolar decision making and change the user’s future clicks or even preferences. These were never certain in the first place, unless the user knows exactly his purpose. Hence, in this paper, we take advantage of such knowledge of the bipolar decision making and hesitation brain indecision in order to propose a serendipity engineering model for learning insights anchored in big and small data methods, tools and analytics in regard to informal learning via social media.

Serendipity is a tendency for making fortunate discoveries while looking for something unrelated; therefore, there is a surprise element attached to it that differs it from the novelty feature. Iquinta and colleagues [3] suggested that there are some context in which user requires unsearched but still useful items or pieces of information. He proposed serendipity heuristics in order to mitigate the overspecialization and accuracy problem with the surprise suggestion.

Changing the mind in a changing environment requires both a time-series research approach for predicting the average user responses as well as qualitative case study to shed light into the process itself. Even though there is mass data analysis at the moment, there is no evidence of improvement into the decision making processes other than improving the systems usability that enable users to reach their targets in a few clicks. ‘Big data’, data so large and complex that processing them by conventional data processing applications isn’t possible or even conventional data collection and analysis, convey information that for some reason is numerical and thus limited. As for education, there was no effort to improve Education significantly. Both ‘Big and small data’ collection and analysis are required to identify suitable and appropriate indicators about bipolar decision making connected to teaching and learning processes, student levels, even social trends.

Education policymakers around the world are now reforming their education systems through correlations based on Big Data from their own national and international education data bases without adequately understanding the details that make a difference in learning paths and processes. Both big data and small data analysis can improve teaching and learning and also indicate or even manipulate trends in education or any other area one desires. Users also act as learners as they need to make decisions under uncertainty or conflicting evidence. Serendipity sequences fill the structural holes in indecision and bipolar decision making, and therefore, aid in proactive decision making and even change of opinion. This is possible via Big and small data analysis, integration and visualisation, cascading and escalating effects in social networks and media [2].

Social Media provide the advantage that are dynamic user generated context, social reactors and mimic groups or even critical masses. In such informal learning environments, causal collisions indicators can serve as predictors for future behaviours. When utilising Chrono-Spatial Intelligent analytics the user can reflect his process on a meta-cognitive level and thus, expand the hesitation gap in order to make better decisions or change initial decisions.

Learning Analytics (LA) is an emerging fast growing area, according to which Learning analytics is the measurement, collection, analysis and reporting of data about learners and their contexts, for purposes of understanding and optimising learning and the environments in which it occurs. Analytics as data mining can be used for teaching, learning and assessing student progress; in this proposition, combining it with chrono-spatial intelligence analytics so to reveal serendipity events for the purpose of learning. Kohavi and Provost [4] suggest five desiderata for success in data mining applications:

  • data rich with descriptions to enable search for patterns beyond simple correlations

  • large volume of data to allow for building reliable models

  • controlled and reliable (automated) data collection

  • the ability to evaluate results

  • ease of integration with existing processes to build systems that can effectively take advantage of the mined knowledge.

There are several LA tools that meet such desiderata, such as: Mixpanel analytics, which offers real-time data visualization documenting how users are engaging with material on a website; Userfly for usability testing and analysis; Gephi is open source for interactive visualization and exploration platform for exploratory data analysis; Socrato for diagnostic and performance reports; SNAPP (Social Networks Adapting Pedagogical Practice) for Learning Management Systems (LMS) data and analysis that visualizes how students interact with discussion forum posts, giving significance to the socio-constructivist activities of students. LA toolkits also already exist, such as the JavaScript InfoVis Toolkit, (http://thejit.org/demos/), Prefuse Visualization Toolkit (http://prefuse.org/) or the learning analytics tool LOCO-Analyst (Liaqat Ali et al. 2012).

Other studies are related to temporal serendipity heuristics and temporal novelty for dynamic features of recommender systems [5], shuffling algorithms [6], through ambient intelligence and interactive datamining [7]. Rana’s research on serendipity for recommender systems focuses on techniques that could predict user interest and assist user’s interaction in finding and following relevant information. Rana is also anchored in making sense past data to accurately predict in future choices. As such, novelty and serendipity refers to the search of finding something new by a user, in our case while searching social media for learning purposes. In order to keep the tricky combination of both accuracy and serendipity novelty, novelty was defined as providing something new which the user have not accesses before but similar in taste while serendipity is a chance discovery that could be really beneficial for a user at certain times and specific purposes. Rana also anchored his research and methodology based on temporal parameters to include the novelty and serendipity in a recommender system. Real-time dynamics is an essential element and play an important role in today’s fast paced life where choices of users depend on many factors susceptible to change anytime [8].

As before, Leong and colleagues [6] conducted a research study analysing online data about the shuffle experience. The results revealed a range of rich and unusual user-experiences and serendipity in particular. Although serendipity is often imbued with ‘magic’ or regarded as a product of chance and luck, its effects can be inspirational and transformative.

Our research on Chrono-Spatial Intelligence and Serendipity contexts [2] are based on global events, people, time and locations can generate visible pathways and connections via Chrono-Spatial Intelligence Analytics. These are: Chrono-Spatial Intelligence Analytics design methodologies as with time-series design; quantitative and qualitative mainstream methods, focus groups and interviews; data analysis via sequential analysis, Natural Language Processing and Social Network Analysis. Indicated Chrono-Spatial Intelligence Analytics Design Methodologies is real-Time series design with streaming data real-time visualisation. Big and small data analysis, therefore, derives from diverse tools, methods and techniques; triangulation is not enough. The suggested research methodologies categories are the following: (a) Spatial Analysis (space – locations); Sequential Data Analysis (events – contexts); Social Network Analysis; and Natural Language Processing. Relationships between big and small data information, events, people, time, locations can be identified by Chrono-Spatial Intelligence Analytics. Context-aware computing can now integrate the suggested methods and techniques in social media platforms translating theories into tools. Time-series real-time streaming data visualisations are set for predefined patterns and peak point identifications, abnormalities identification, and decision making identification points.

Context-aware time is also related to presence tools integration into social media. A user’s presence information includes audio and visual cues for the following factors: self-presentation, distance, location, activity, feeling, form of expression, availability, willingness and readiness for communication, current device or application on which he can be communicated (cell-phone, home & work-hub landline phone, IM client), motivation and enjoyment, prediction. Presence brings this important information to the social network incorporating sets of different types of information that enables better communication and range of feelings. Deep presence and copresence enable embodied communication and immersion with user’s situation (feeling and activity of the user, buddies, friends and broader networks) that possibly influences the main user to act or silent accordingly. Presence is also a critical element is serendipity engineering related to bipolar decision making on increasing awareness regarding the learner learning paths. Time-based coordination examines the nature of decision, clicking or not clicking on the suggested information item: (a) the form of the effect (the level, slop, variance and cyclicity); (b) its permanence (continuous or discontinuous) and (c) its immediacy (immediate or delayed). The research question now is revolved around the ways such serendipity events can be integrated into a learning system. This is discussed next.

3 To Click or not to Click? Brain Hesitation and Randomness

People use social media from different education levels, cultural backgrounds, and ages can provide us with the underpinned structures, relations, activities, scale and impact of such networks and human relationships [9]. Therefore, we may create a model to identify and measure cues as well as users’ behaviour for learning purposes, as for example user’s levels of engagement in serendipity events and the way they learn to change their decision for greater learning impact. In this way, we may possibly predict serendipity as happy accidental surprises or even engineer it. Data can be extracted and analysed via specific and valid methods, tools and techniques mainly focusing on users’ preferences, experiences and satisfaction. Another serendipity element is the attached value and importance to the related person; a serendipity event may be important with emotional arousal, capture of attention and immediate response to engage and click for one person and completely insignificant and seemingly unimportant for another. Taking this concept further, a person or/and application has to be proactive for such serendipity events to occur. This may be a behavioristic approach in education; however, utilising the creative euphoria for learning is not only for marketing purposes but for enhancing educational behaviours as well via repetitive reinforcement towards the educational target.

Experiencing serendipity is a pleasant surprise and such it conveys intrinsic satisfaction of being lucky. Perceiving serendipity is also a decision making challenge as the same information may appear relevant or irrelevant to the user. As such, we argue, that the levels of emotional achievement and satisfaction may aid in changing serendipity perception and insightful creative thinking via social media. Understanding the way that people think and make associations among their own interests, resources and other people can encourage serendipity with even more communication and collaboration towards serendipity.

Randomness is an occurrence with no definite aim or guidance in a particular direction and without any method or conscious choice (adapted from the Oxford English Dictionary on ‘random’). In systems design, a large number of alternatives can be randomly generated, which are then reduced to a smaller set by a non-random selection process. A number of creative ideas can randomly be triggered and then assimilated into a pre-existing system of beliefs in order to sustain its consistency and integrity, which is a non-random process of reinforcement [11].

Unexpected associations are influenced by in situ chrono-special intelligence i.e. they are context-aware relations [2]. Such relations indicate information organisational convergence, coordination and sequencing. Coordinating and sequencing learning activities in social media is of major importance and contradicts educational material fixation. The degree of coercion and tackling unpredictability is a delicate design choice [10]. Therefore, predicting and assists such learning sequences can scaffold students’ social and learning interactions in such a way so to be taken by learning surprise and achieve peak performance based on excitement and creativity. Prediction of specific linear and non-linear sequences online for preferable activities and resources as well as repetition of such actions and clicks with minor variations can be used. Traversion and rotation are two modes to be used. The first is the repetition of the same educational material in the same order looped through as extracted from learning analytics, with only one element being in use at a time; the second is rotates the elements in a given set towards the same direction.

Randomness in this sense can be used as an innovative learning serendipity design as a resource for supporting rich and novel user experiences when navigating in social media. Leong and his colleagues [6] assert that encounters with serendipity are more likely to arise when the random selection is unconstrained (i.e., when the shuffle algorithm picks tracks randomly and freely from people’s entire music or photo library instead of being constrained to only subsets of the library).

4 Discussion: Taking Learning by Surprise via Serendipity Engineering in Social Networks

In our research proposition, social and emotional needs in design refer to safety, social and esteem needs that can trigger the click i.e. engagement and actualisation. The research design follows our previous research on Chrono-Spatial Intelligence Analytics [2] for (a) Spatial Analysis (space – locations); Sequential Data Analysis (events – contexts); Social Network Analysis; and Natural Language Processing. Other researchers such as Rana and colleagues [4, 7] conducted studies on web personalization, customizing recommendations towards the needs of each specific user or set of users, and taking advantage of the knowledge acquired through the analysis of the user’s navigational behavior. Profiling [2] and data analysis integration updating the learner’s profile can alleviate the information overload problem so social media can be friendlier and more easy to use. As each user has also the learner personal on social media and the Internet in general, experiences serendipity differently, profiling and personalisation can solve several usability and pedagogical usability overloading problems. Such challenge for social media designers requires involving serendipity engineering into the initial systems design so to include learning analytics together with business analytics for both past analysis and future prediction. In other words, analytics can also be used for learning purposes not just marketing and business.

Novelty and serendipity has been investigated together to support human-centered discovery of knowledge; however, shuffling novelty in general and new learning suffers due to lack of related research. Social media mining for learning purposes needs to be optimised towards serendipity and novelty in learning using temporal dimension. Serendipity Engineering in social media builds upon randomness and interaction, relaying on user’s bipolar decision making processes and hesitation before each click, so to embrace innovative learning in favour of the user as learner. In this way, serendipity and unintended outcomes can be manipulated to orchestrate pleasant learning surprising experiences.

Hidden and unexpected connections for Serendipity Engineering related to the bipolar decision making process and randomness can work within the decision making hesitation gap existing before taking the risk on a click. Further action depicts the user’s acceptance and engagement in serendipity or even denial to participate. The proposed serendipity engineering model based on both engineered learning insights and random learning items and events are anchored in big and small data methods, tools and learning analytics.

5 Concluding Remarks and Future Work

This paper discussed the possibility of including learning analytics together with business analytics in social media in order to support serendipity in a user’s learning path. This notion is supported by the ways the human brain works during the decision making processes. When shuffling online for information items or else, there is a hesitation gap before entering an action such as a click so to engage in a series of events, hopefully ending in a pleasant and magical surprise and experience. Such magical immersive experience has enormous learning gains if taken advance effectively by the user. The user in online and social networks acts also as a learner.

Fortunate and unfortunate series of events can lead the user by preference or avoidance to act with a click. Such underpinned matrix can be investigated to reveal hidden connections base upon Chrono-Spatial Intelligence for Serendipity Engineering (CSISE). Digital CSISE depends on the effectiveness for past behaviour analytics combined with user profiling and preferences between human or machine generated data. Social media data in correlations with other data formats can provide information for real-time political impact related to political decision making, as for example quality in education. Social media also provide by definition important temporal information related to user’s presence and rich context interactions.

In this paper, only the individual approach for Chrono-Spatial Intelligence for Serendipity Engineering (CSISE) for learning purposes in social media was discussed. Groups and community usage and collaborative human factors were not the proposition target. Nonetheless, social interactions for group activities can provide a mass moving serendipity effect which is not the purpose of this paper. Serendipity in learning engages the user as learner in 3 ways: (a) behavioural engagement with involvement and participation; (b) emotional engagement, regarding positive feelings and surprise as a pleasant experience used in learning in general; and (c) cognitive engagement, as conscious efforts for excelling in acquiring comprehensive knowledge and skills when in social media.

In regard to technological factors, CSISE for learning purposes can aid users as learners to reach a meta-cognitive awareness on past actions in order to enrich their preferences and online engagement in suitable and appropriate learning paths for achieving excellence. Furthermore, users as learners can also be aware of their constraints and disadvantages and build upon reversing difficulties. Traditional tools and models of learning engagement are more teacher-oriented and are designed based on complete scripting of the learning behaviours and educational material, omitting the serendipity element.

Lastly, in regard to research methodologies and suggested CSISE learning analytics for social media as such, both quantitative and qualitative results as well as other sources of data visualisation and utilisation need to be converged for a high level of engagement in predicted serendipity events, resulting in high level of learning and understanding of online information. Utilising a range of tools, methodologies and techniques, can create intervention and interference based on prediction can actually alter the future decision making users’ domino of events and items selected and orchestrate time-based series of events that may have possible utility in psychology, marketing, learning and business analytics, serendipity systems design and engineering design or other influential disciplines.

To conclude, the random presentation of objects can engage or intrigue people; however, in combination with serendipity past activities research can aid people to attract their attention and drawing them into the serendipity flow. A change of mind may also imply a change into a new string of possible events and items to click. There is only commitment and determination that diminish the hesitation moment [12]. Such drawing of attention can also be part of a thought suspension process or defamiliarization; introducing a slightly strange item or a gap can also heighten attention Serendipity based on randomness and unpredictability is an important and exciting element and social media systems feature for users’ immersive experience for to inform CSISE subsequent design and application.

In our future research, sentiment text analysis for positive, negative or neutral attitudes including the indecision hesitation gaps follows our research stream on Chrono-Spatial Intelligence for Serendipity Engineering (CSISE) for learning purposes. Revealing any interaction patterns utilising past data, in fact reveals learning attitudes and behaviours for the future. Changing a deterministic and fixed future may be of interest for anyone involved in learning and human excellence.