Abstract
Twitter is currently one of the most imperative and powerful social media platforms. Besides, Twitter has become one of the significant new channels for public communications. The studies on the uses of Twitter have emphasized on the isolated and critical events including political and crisis communications, popular culture and events such as sports.
The studies also center on the applicability of the Twitter’s search stream APIs, which often provide a coordinated keywords prefixed by hashtag ‘#’. Twitter provides a critical degree of data access through its Application Programming Interphase (API) platform. The API is an interface that is designed to be used by Twitter clients.
However, the interface can also be applied in tracking current events by other users especially researchers through the use of a specific keyword and hashtag. The Twitter API allows researchers and other users to track and inevitably apprehend tweets with particular hashtags or keywords.
Introduction
Twitter is currently one of the most imperative and powerful social media platforms. Besides, Twitter has become one of the significant new channels for public communications. The studies on the uses of Twitter have emphasized on the isolated and critical events including political and crisis communications, popular culture and events such as sports.
The studies also center on the applicability of the Twitter search stream API, which often provide a coordinated keywords prefixed by hashtags ‘#’. Researchers have used the Twitter’s API platforms to access, analyze and store data. The API is an interface that has been designed to be primarily used by Twitter clients.
However, the interface can also be applied in tracking current events by other users especially researchers by customizing specific keyword and hashtags. The Twitter API allows researchers and other users to track and inevitably apprehend tweets with particular hashtags or keywords.
In an advanced usage, the API enables third party developers attach novel add-ons to the prevailing services (Lomborg & Bechmann, 2014). The API provides interface that allow researchers to gather more information off a particular social media services that can be used for quantitative or qualitative analysis (Lomborg & Bechmann, 2014).
In addition, the API provides the real-time data within the social media network such as Twitter (Purohit, Hampton, Shalin, Sheth, Flach & Bhatt, 2013). As such, Twitter’s open APIs offers unprecedented opportunity for physicist, linguists, computer and social scientists to study the behaviors of people within the social media platform (Kwak, Lee, Park & Moon, 2010).
In fact, researchers have utilized the Twitter API to study varied human behavior across various disciplines ranging from business management to crisis response, management and communications (Marwick, 2012). The Twitter’s API has been applied by researchers to study various global events. For instance, Bruns, Highfield and Burgess (2013) used the API platform to study the Arab Spring.
The API as a Data Collection Tool
As indicated, Twitter provides a critical degree of data access through its Application Programming Interphase (API) platform. The APIs platform offered by Twitter allows researchers and other users to track and inevitably apprehend tweets with particular hashtags or keywords (Bruns & Burgess, 2012).
The increased accessibility of Twitter’s information through the use of its API platform has enabled researchers to conduct studies on various aspects of human behavior patterns. Increased novel applications of Twitter’s API has been developed recently (Naaman, Boase & Lai, 2010).
In fact, Twitter has allowed researchers in the fields of healthcare simulations as well as in education to share most relevant information through the application of specialized tools within the API platform. The capability has enabled the fields to conduct studies on various fast evolving events. Such kinds of studies are made possible due to the simplicity of the Twitter’s API (Shi, Rui & Whinston, 2014).
Given the simplicity of the Twitter’s API, the popularity of diverse web-page application has increased (Bagley, 2012). Such web-page applications have provided value-added services to the Twitter users. For instance, Twuffer web application allows Twitter users to schedule tweets for future dates.
Twidentify is also another web-based Twitter search engine that allows users to examine tweets through the use of keywords. Such tools are normally applied in the API platform and enables researchers to easily follow trends on a particular event through the application of keywords or certain words prefixed by the hashtag (Humphreys, Gill & Krishnamurthy, 2014).
The Twitter API uses three ways of searching through the application of keywords. One of the methods is the trend search, which allows researchers to track the popularity of the keywords over a particular time (Yardi & Boyd, 2010). The second method is the basic Twitter search that tends to track people who utilize the keywords on their tweets and re-tweets.
The final method is the search on influence. The data from the searches are normally sorted in tweets and re-tweets. Such information is provided in diverse formats including time series graphs. The trend search, basic Twitter search and the search on influence form the foundation of various models researchers have utilized to study innumerable behaviors and events patterns on Twitter (Humphreys, 2011).
Researchers utilize various tools to collect and analyze the information, which are then customized into a single timeframe and serve various functions including comparisons to the statistics of different studies.
The Opportunities and Challenges in the Application of APIs
Lomborg and Bechmann (2014) argued that APIs enables social media researchers to have an easy access of data particularly from the social media companies. Essentially, the APIs enables third party developers attach novel add-ons to the prevailing services. The APIs provide interface that allow researchers to gather more information off a particular social media services that can be used for quantitative or qualitative analysis (Lomborg & Bechmann, 2014).
APIs as a research tool offers significant opportunities for the studies that are either qualitative or quantitative. The API platform allows easy automation of data collection, analysis and storage (Lomborg & Bechmann, 2014). The cleaning and storage of data are also conducted sequentially and automatically.
Combined with other web-based digital research tools, the data accessibility are traced to their usage patterns in greater detail more than what other data collection methods such as interviews and surveys provide. Through the use of API, the data collection is prompt and nonintrusive (Lomborg & Bechmann, 2014).
How Various Researchers have used API as the Data Collection Method
The API platform has been used widely by various researchers to study a wide range of correlations on particular sub-sets of human behaviors and events patterns within Twitter. For instance, Boyd and Crawford (2012) examined how large volumes of data can be mined, evaluated and stored through the application of API tools.
Misopoulos, Mitic, Kapoulas and Karapiperis (2014) applied the Twitter API tools to examine the trends in customer behaviors and experiences in the airline industry. Further, Mooney, Winstanley and Corcoran (2011) also examined how API offered by Twitter can be used to evaluate environmental issues.
Moreover, Bruns and Burgess (2012) examined innovative approaches applied by researchers to collect and analyze large quantity of data from Twitter. In their study, the researchers outlined the manner in which Twitter is applied in covering the main events with emphasis on the use of hashtags, which are the keywords that identify a tweet as critical in an ongoing discussion.
Twitter provides a critical degree of data access through its Application Programming Interphase (API) platform. While API is an interface that is designed to be predominantly used by Twitter clients, it can also be applied in tracking current events by other users (Bruns & Burgess, 2012). The limitations on such processes are comparatively low and the tools for such examinations are provided to the researchers.
Researchers can apply various API tools that are readily available including yourTwapperKeeper (yTK) to access data (Bruns & Burgess, 2012). With minor modifications in order to make data export improvements, the yTK is normally applied to track a significant number of keywords on particular events spontaneously.
The increased processing and assessments of the collected data sets discloses numerous significant patterns. Numerous approaches of data analysis are also available to the researchers (Bruns & Burgess, 2012). The network analysis and visualization, statistical packages and data-processing software are available to the researchers.
The researchers would apply these data processing tools depending on the variables under the study as well as the methodology (Bruns & Burgess, 2012).Further, API has been found to provide the real-time data within the social media network. Purohit et al. (2013) applied the Twitter streaming API to provide the real-time information from the tweet collections.
Through the use of the Twitter search API, the researchers are capable of selecting conversational tweets in one response based on the keyword feed preceded by the hashtag. Using the keyword-based method, the Twitter search API is capable of providing a bout 1500 recent tweets in a single response while eliminating tweets from private users. The search query is capable of providing the related metadata including the timestamp, location and the details of the author.
The API can be applied to study the topological characteristics of Twitter and its capability as a novel medium through which people share information. The API can be used to collect information on the user profile, trending topics and the tweet messages. The researchers can use various analytical tools including network, ranking and trending analyses to examine the attributes of Twitter.
The study on the topological characteristics of Twitter was first conducted by Kwak et al. (2010). The study conducted by Kwak et al. (2010) formed the foundation on which other studies using the API as a method of data collection to examine the attributes of other social media networks are based. The Twitter’s API offer unprecedented opportunity for both physical and social scientists to study the behaviors of people within the social media platform (Kwak et al., 2010).
Researchers have also used API to store and analyze data for real-time applications. Kumar, Morstatter and Liu (2013) explored how information from Twitter can be collected, stored and analyzed. In fact, Kumar et al. (2013) examined the collection of Twitter information through its free APIs and the manner in which such information can be stored for real-time applications.
Various measures and algorithms can be applied in the analysis of data sets from various social media networks (Kumar et al., 2013) The free Twitter API is a critical data collection tool that has led to the development and application of studies in diverse areas including humanitarian assistance and disaster relief and management.
Disaster relief and management as well as the humanitarian assistance are areas that have been widely benefitted from Twitter information to increase awareness to the crisis situation (Kumar et al., 2013). Through the application of free Twitter APIs, the researchers can accurately predict the occurrences of major disasters such as earthquakes and tsunamis by identifying the relevant users to follow in order to acquire the disaster related information (Kumar et al., 2013).
The Twitter’s API has been widely used to study disasters and major events across the globe. For instance, Murthy and Longwell (2013) applied the Twitter API to examine how users tweeted and re-tweeted during the Pakistan floods. In their study, the researchers examined the tweets and re-tweets links between the traditional and social media and in which countries the most tweets are from as well as relationship between the location and the traditional versus social media.
The use of API normally follows similar procedures. The only differences are the methods of analysis and the subject of the study (Murthy & Longwell, 2013). The studies by Kumar et al. (2013) as well as Murthy and Longwell (2013) provided instances where the researches on the applicability of social media in relation to disaster events have grown over the past years covering a wide range of sites and procedures.
The ongoing researches such as those conducted by Kumar et al. (2013) as well as Murthy and Longwell (2013) on the use of API have centered on Twitter use as a critical turning point in increasing awareness as well as a measure through which such problems can be mitigated. The uses of Twitter in covering most of the disastrous stories legitimize the site as a journalistic space ((Murthy & Longwell, 2013).
As such, researchers including Kumar et al. (2013) as well as Murthy and Longwell (2013) have concentrated their studies on how users tend to tweet and re-tweet in order to examine the trends, which remain critical information in predicting the major global events. The volume of data also affects the manner API is used in the studies. For instance, Murthy and Longwell (2013) collected their Twitter data through the accessibility of Twitter search API stream during the time when the text string “Pakistan” was most active.
Given the large volume of data, the researchers applied the Social Network Analysis (SNA) to scrutinize the data and come up with the relationships under the study. Through various analytical models researchers can come up with varied patterns on data.The ongoing application of Twitter API on main events has emphasized on single cases (Bruns & Stieglitz, 2012).
The comparative studies on the dynamics of patterns on numerous events are deficient. Few comparative studies have resulted into new discrete discussions. For instance, in their comparative study on the dynamics and patterns of large number communication events on Twitter, Bruns and Stieglitz (2012) identified various discrete forms of deliberations that can be witnessed on Twitter.
Through the use of a wide range of communication metrics, Bruns and Stieglitz (2012) indicated that diverse types of tools available to Twitter users can be influenced by thematic and contextual factors. The study by Bruns and Stieglitz (2012) is an example of an analytical view of the inclusive metrics of the Twitter discussions from various contents.
Further, the studies on the uses of Twitter through the application of its API have emphasized on the isolated and critical events including political and crisis communications, popular culture as well as events such as sports (Bruns & Stieglitz, 2013). The studies also center on the applicability of the Twitter search stream API, which often provide a coordinated keywords with hashtag ‘#’ placed in front of the word (Bruns & Stieglitz, 2013).
However, such studies are deficient of standard metrics that can be used to relate message arrangements across the cases and events under the study (Bruns & Stieglitz, 2013). While addressing these problems of the API methodology in their study, Bruns and Stieglitz (2013) provided an index of extensively applied homogeneous metrics for scrutinizing Twitter-based communications with central focus on exchanges that emphasize a particular hashtag.
The researchers presented the importance of such indices in comparing communication patterns and the extent to which the use of such metrics can attain in comparing communication exchanges. While most Twitter studies emphasize on the detection of events through the application of algorithms that depends on keywords and the capacity of tweets on that particular events, an alternative analysis of tweets also occurs (Graves, McDonald & Goggins, 2014).
Graves et al. (2014) looked at the occurrence of alternative analysis through the use of low-volume data. Instead of using high-density tweets, the researchers analyzed low-volume tweets based on keywords though not linked to spikes in tweets-per-minute. The events Graves et al. (2014) studied are not readily detected by the current API event-detection algorithm, which depend on the tweet volume to drive the analytical engine. The study by Graves et al. (2014) demonstrated that there are various ways in which Twitter can be conceptualized through the application of theories integrated within the study.
Challenges and Limitations on the use of Twitter API
The applicability of API as a method of data collection has several limitations (Metzger, 2014). API is not only acting as a barrier in information gathering due to privacy precautions, but also in terms of the applicable tools (Metzger, 2014). Bruns and Stieglitz (2012) indicated that thematic and contextual factors influence the application of the tools available to the Twitter’s API users.
The limitations affect the academic researchers that wholly depend on tracking the current events and activities as well as thematic interests of large and representative samples of Twitter users (Bruns & Burgess, 2012). The use of the digital footprint in the data collection, analysis and storage also present various issues (Lomborg & Bechmann, 2014). In other words, applying web-based tools combined with APIs platform is deficient in terms of generalization of data. Researchers have always resented on the generalization of data through the use of digital platforms.
Lomborg and Bechmann (2014) presented significant procedural deliberations of the prospects and problems linked to both qualitative and empirical social media studies through the application of the APIs. Generally, wide-ranging methodological issues should be taken into consideration while gathering and evaluating data through the application of the APIs.
Further, it is critical to examine the effects of legal and moral implications of empirical studies that apply APIs as a data collection method (Lomborg & Bechmann, 2014). Issues related to generalization, reliability and validity of data collected through the use of APIs are often emerging. The use of content is also vcevolving as a critical legal and ethical issue while applying APIs as a data collection method in the empirical studies (Lomborg & Bechmann, 2014).
Moreover, the information available through API is partial unless one is a third-party API provider. The restrictions on the use of APIs affect the academic studies that wholly depend on tracking the current events and activities as well as thematic interests of large and representative samples of Twitter users (Bruns & Burgess, 2012). The applications of tools are also influenced by various factors including cost constraints, security and accessibility.
While information from Twitter can easily be obtained through APIs, the API as a data collection method has various challenges that must be observed. According to Kumar et al. (2013), the Twitter streaming API provides about 1% of the sampled Twitter data. As such, care should be taken while conducting studies from information collected from the streaming Twitter API.
Conclusion
The studies of events and human behaviors on Twitter through the application of APIs have taken various forms and procedures. The APIs enables social media researchers to have an easy access to data particularly from the social media companies such as Twitter and Facebook. In essence, APIs interface enables third-party users to attach novel add-ons to the prevailing services.
The APIs provides interface that allow researchers to gather more information off a particular social media services that can be used for quantitative or qualitative analysis. In this discussion, it has been shown that the research tool has been applied to study events that range from the crisis to business management on Twitter. However, the use of the tool has various deficiencies that should be considered while conducting the studies on the social network.
References
Bagley, C. A. (2012). Chapter 6 embracing Twitter as a research tool. Library and Information Science, 5, 131-147.
Boyd, D. & Crawford, K. (2012). Critical questions for big data. Information, Communication & Society, 15(5), 662-679.
Bruns, A. & Burgess, J. (2012). Researching news discussion on Twitter. Journalism Studies, 13(5-6), 801-814.
Bruns, A. & Stieglitz, S. (2012). Quantitative approaches to comparing communication patterns on Twitter. Journal of Technology in Human Services, 30(3-4), 160-185.
Bruns, A. & Stieglitz, S. (2013) Towards more systematic Twitter analysis: metrics for tweeting activities. International Journal of Social Research Methodology, 16(2), 91-108.
Bruns, A., Highfield, T. & Burgess, J. (2013). The Arab Spring and its social media audiences: English and Arabic Twitter users and their networks. Queensland, Australia: ARC Centre of Excellence for Creative Industries and Innovation, Queensland University of Technology.
Graves, I., McDonald, N. & Goggins, S. P. (2014). Sifting signal from noise: A new perspective on the meaning of tweets about the “big game”. New Media & Society, 1(1), 1-20.
Humphreys, L. (2011). Who’s watching whom? A study of interactive technology and surveillance. Journal of Communication, 61, 575–595.
Humphreys, L., Gill, P. & Krishnamurthy, B. (2014). Twitter: a content analysis of personal information. Information, Communication & Society, 17(7), 843-857.
Kumar, S., Morstatter, F. & Liu. H. (2013). Twitter data analytics. New York, NY: Springer
Kwak, H., Lee, C., Park, H. & Moon, S. (2010). What is Twitter, a social network or a news media? World Wide Web Conference Committee, Raleigh, North Carolina, USA.
Lomborg, S. & Bechmann, A. (2014). Using APIs for data collection on social media. The Information Society: An International Journal, 30(4), 256-265.
Marwick, A. E. (2012). The public domain: Social surveillance in everyday life. Surveillance and Society, 9(4), 378–393.
Metzger, M. (2014). Privacy, trust, and disclosure: Exploring barriers to electronic commerce. Journal of Computer Mediated Communication, 9(4), 335–361.
Misopoulos, F., Mitic, M., Kapoulas, A. & Karapiperis, C. (2014).Uncovering customer service experiences with Twitter: the case of airline industry. Management Decision, 52(4), 705-723.
Mooney, P., Winstanley. A. C. & Corcoran, P. (2011). Evaluating Twitter for use in environmental awareness campaigns. Web.
Murthy, D. & Longwell, S. A. (2013). Twitter and disasters. Information, Communication & Society, 16(6), 837-855.
Naaman, M., Boase, J. & Lai, C. (2010). Is it really about me? Message content in social awareness streams. New Brunswick, NJ: Rutgers University, School of Communication and Information.
Purohit, H., Hampton, A., Shalin, V. L., Sheth, A. P., Flach, J. & Bhatt, S. (2013). What kind of #conversation is Twitter? Mining #psycholinguistic cues for emergency coordination. Computers in Human Behavior, 29, 2438–2447.
Shi, Z., Rui, H. & Whinston, A. B. (2014). Content sharing in a social broadcasting environment: evidence from Twitter. MIS Quarterly, 38(1), 123-142.
Yardi, S. & Boyd, D. (2010). Dynamic debates: an analysis of group polarization over time on Twitter. Bulletin of Science Technology & Society, 30(5), 316–327.