THE EFFECTIVENESS OF WEB-BASED TECHNOLOGY PLATFORMS IN FACILITATING CONSTRUCTION PROJECT COLLABORATION: A QUALITATIVE ANALYSIS OF 1,152 USER REVIEWS

SUMMARY: The construction industry accounts for 9% of global GDP. Efforts at addressing construction’s inherent inefficiencies have over the last decade increasingly involved the deployment of web-based collaborative tools. Consequently, much research has been devoted to assessing these platforms; including interoperability, workflow management and technological limits. What has not been considered to date are the views of web-based tool users themselves as to the functionality, potency and usability of the various platforms available on the market. Currently, there are 5,300,000 documented users of web-based collaborative tools. If web-based collaboration is to be further enhanced, the views of users must be known. This study explores this dimension. Financeonline’s top six tools were considered: CoCostruct, PlanGrid, Autodesk BIM 360, Procore, e-builder and Aconex. Around 200 reviews for each tool were collected from ‘Business Software Reviews from Software Advice,’ resulting in a total dataset of 1,152 complete reviews. Text-mining analysis was applied to this dataset, using RapidMiner Studio 7.5. Thirty key terms with a frequency of over 100 occurrences were retrieved; terms such as software, manage, inform, support, easy use, function, track and friendly. These constitute the subject of the reviews. These terms were then analyzed for sentiment qualifiers; either positive or negative. A total of 804 sentiments were positive, 322 negative and 26 neutral. This study thus highlights that while 70% of user reviews of web-based collaborative tools are positive, there remains much room for improvement. Areas for improvement are also indicated by this study.


INTRODUCTION
The construction industry accounts for 9% of Global GDP (McKinsey & Company, 2017), having an added value of $3.6 trillion (US dollars) and annual revenue of almost $10 trillion, which is estimated to generate revenue up to $15 trillion by 2025 (Forum, 2016;McKinsey & Company, 2017;StartupAUS, 2017). Despite its significance, the construction industry is affected by poor design information coordination and inefficient communication that give rise to project delays, change orders and conflicts (Bouchlaghem, 2012). These problems are further exacerbated by the sector's highly fragmented nature (Boton & Forgues, 2017), uniqueness of each project and unstructured working conditions. As a remedial solution, technological innovations are adopted which can result in a 1% productivity rise worldwide and save $100 billion (Forum, 2016;McKinsey & Company, 2017;StartupAUS, 2017). Chief amongst a plethora of technological developments, Building Information Modeling (BIM) is defined as a modeling process and connected set of procedures to yield, link and analyze data-rich models. Palpable benefits of BIM include: better-quality planning, improved design, economical construction, and better operation and maintenance processes. When using BIM, one or more virtual computer models for construction are shaped digitally, that contain detailed geometry and rich data required to support the construction. BIM, as an emerging technological innovation, relies on smooth collaboration among project team members (Merschbrock & Munkvold, 2015). BIM-enabled projects are heavily reliant upon collaboration tools, as knowledge management or digital support technologies for workflow management and data exchange. A BIM presenting model server is likely to simplify the exchange of data in a multi-model situation. This is achieved by supporting the numerous applications involved in a building project's life-cycle plus design tools, analysis tools, facility management tools, electronic document management systems (EDMS) etc. (Shafiq et al., 2013;Singh et al., 2011). EDMS cf (Edwards et al., 1996) and web-based project management applications are another form of collaboration technology. These aforementioned packages are considered as collaboration tools in this study. As technological developments exponentially increase, collaboration technologies, tools and processes enhance productivity and reliability as they are supporting construction process, management and collaboration (Hardin & McCool, 2015).
Evidence shows that web-based collaboration tools and associated mobile application implementation are rising and 72% of US construction professionals are utilizing smartphones at work (O'Malley, 2015). The Associated General Contractors of America (AGC) and Sage Group 2016 Business Outlook Survey indicated that 63% of construction businesses are implementing cloud-based platforms to improve information access from a different location (Kracunas & Wetmore, 2016). Anecdotal evidence suggests that this implementation figure is set to increase. The number of existing collaboration tools varies according to different sources from 325, 349 or 374 tools (Capterra, 2018;Crowd, 2018;Softwareadvice, 2018). Along with the number of tools, the ranking of the best/ top tools also varies according to different sources thus compounding a dilemma of selecting an appropriate tool. Moreover, the inconsistency of tools' ratings and variation of internal methodology results in different decision outcomes.
Previous studies have focused on the: effectiveness of collaboration tools (East et al., 2008); technological aspects (Zhang et al., 2017) ; identification of barriers and frameworks for improvement of collaboration and communication (Mignone et al., 2016); and communication between project teams (Hosseini et al., 2017). In addition, teamwork improvements via utilizing collaboration tools (Chung et al., 2009;Costa & Tavares, 2012) have also been explored, along with the identification of the antecedents and drivers of collaboration technology adoption. This body of work has hitherto either relied on individual case studies or on informal evidence provided by 'successful stories' reported in the trade press. In these studies, the project team faced various individual, environmental and technological challenges while implementing digital processes (Merschbrock & Munkvold, 2015). There are many factors that may repress knowledge sharing throughout industry. Although the former studies identified the barriers and challenges of collaboration tools, the collaboration tools users' perceptions were not considered in detail. Consequently, this research aims to address this gap, through: analyzing the reviews of six prominent collaboration tools users' reviews; identifying the positive and negative aspects in general; and revealing the frequently appeared aspects of the tools. The study contributes to the prevailing body of knowledge by providing a cross sectional snapshot picture of existing users' perceptions, which will be invaluable in assisting vendors and system developers as well as digital managers and project leaders of construction projects who strive to augment performance.

Collaboration in the construction industry
According to Wood and Gray (1991, p. 146): "Collaboration occurs when a group of autonomous stakeholders of a problem domain engage in an interactive process, using shared rules, norms and structures, to act or decide on issues related to that domain.'' To fulfill common objectives, collaboration occurs within an atmosphere of trust, openness and honesty by several individuals who undertake a process of sharing collective knowledge, expertise and skills (Mignone et al., 2016). In the construction industry, multidisciplinary collaboration is a key success factor for all the parties involved in delivering projects (Singh et al., 2011;Zhang et al., 2017). Effective collaboration among parties is required to ensure mutually beneficial relationships among parties to jointly create appropriate rules and structures (Oraee et al., 2017). Collaboration allows geographically dispersed project teams to work together to: increase efficiency in the process collectively; provide greater profitability to the organization (S. Moses et al., 2008); and enhance the performance of the construction sector (Comiskey et al., 2017).
That said, collaboration in the construction industry is challenging due to a wide range of reasons and lack of collaboration engenders numerous continuity problems (Oraee et al., 2019). Specifically, many stakeholder organizations within the sector are stagnant, thus leading to highly fragmented industry with low productivity and dominated by small businesses (Costa & Tavares, 2012). Effective collaboration can overcome these challenges by integrating among design and production processes cf. (Ahsan et al., 2007;Bi et al., 2019).Failure to do so results in time and cost overruns, poor coordination, less than optimum information sharing, and inadequate, inappropriate and inconsistent communication (Durdyev & Hosseini, 2018). Moreover, lack of collaboration results in a proliferation of adversarial relationships among project stakeholders. Misunderstandings, misinterpretations of data and increased rework may result in project delays attributed to ineffective collaboration practices (Mignone et al., 2016). In short, collaboration is quintessentially important in ensuring efficient and effective construction procedures (Costa & Tavares, 2012) and it is indispensable to the success of construction projects (East et al., 2008;Mignone et al., 2016).

Construction collaboration tools
The proliferation of Information Technology (IT), and the advent of web-based applications in construction activities profoundly transformed the collaboration in construction industries in recent years (Oraee et al., 2017). Moreover, Information communication technology (ICT) development and maturity resulted in an increasing trend of transferring activities from offline to online (Ma et al., 2018) and enabled standardized communication between different actors in Construction (Adriaanse et al., 2010;Hosseini et al., 2017).
Globally, construction management software is forecasted to grow at a compound annual growth rate (CAGR) of 9.19% between 2018 and 2022 (Markets, 2019).Consequently, Data exchange over the web is gaining popularity within the construction industry (Anna Wagner et al., 2020). Likewise, with the emergence of cloud-based technologies, many connected job sites can transfer and make available every aspect of project information to all the relevant parties anywhere in the world (Deloitte, 2020). For contemporary projects, Computer-based collaboration has become the standard for scattered team members across different locations (Oraee et al., 2017). Furthermore, Incorporation of integrated BIM modules assembled in hybrid platform support collaborative web tools and BIM server resulted in a collaborative working environment (Charalambous et al., 2017;Costa & Tavares, 2012).
Along with continuous development in information and communication technology (ICT), projects are increasingly more complex and involve larger capital investments, dispersed project participants, and tighter schedules . Consequently, Virtual meetings, tele-and audio-conferencing technology, instant messages, 3D, virtual and mixed reality are considered synchronous collaboration platform and have become the norm for contemporary projects Ma et al., 2018;Oraee et al., 2017).
In short online collaboration, platforms can be denoted as the amalgamation of web-based technologies which offer a shared interface by linking multiple interested parties, and provide a platform in a digital form to share, exchange and store project information and work in collaboration on a basis of subscription fee, license plus maintenance, negotiated fixed cost or exclusive business partnership agreement (Charalambous et al., 2017). According to Adriaanse et al. (2010) "A Digital coordination and collaboration tool used for communicating and sharing project information between participating organizations in construction projects" is defined as an interorganizational ICT. Moreover, Cloud Computing, software-as-a-service (SaaS), and Service-Oriented Architecture (SOA) concepts are also associated with online collaboration platforms (Charalambous et al., 2017). Comiskey et al. (2017), indicated "Common Data Environment (CDE) as an internet-based cloud hosting platform accessible to all construction team members' access shared project information". Collaborative technology also refers to tools and systems specially designed for better facilitating the group work both in the office and remotely.
In this study, web-based construction collaboration tools are denoted to all the online collaboration platforms, digital collaboration and coordination tools, construction management software, Cloud-based Common Data Environment (CDE), Collaborative working environment and so on which altogether enhance collaboration in Construction.

Collaboration tools Benefits
Web-based collaboration tools have different features and applications to assist the coordination of business processes and enable collaborative workflow, data access, multiple team, and office locations integration (Capterra, 2020). Project management and customer management functionality along with accounting, scheduling, and portfolio management services often incorporated in these tools (FinancesOnline, 2020).The benefits of webbased collaboration tools have been discussed in the following segment.

Efficient Document and Project information Management
Document management applications of web-based collaboration tools assisted in storing, organizing, and managing documents in a digital way within construction projects (Adriaanse et al., 2010). Moreover, effective information management and exchange (Comiskey et al., 2017) can be achieved through collaboration tools. BIM and digital technologies foster the integration of activities and strengthen the management of projects  as well. Engineering Project network Team members currently link project members electronically and transfer and process project data to and from disperse participants . Furthermore, document control capabilities of tools improve accountability by ensuring tracking and version control and minimizing liabilities through a virtual paper trail for plans (softwareadvice, 2020). In brief, collaboration tools in construction enhance document and project information management.

Improved Communication and Collaboration
Collocated teams highly dependent on technology (Hosseini et al., 2017) to exchange data and communication between team members. Working as a central repository for project information web-based collaboration tools facilitated the sharing of resources between geographically dispersed teams and enhanced communication (Charalambous et al., 2017). Moreover, Inter-organizational cooperation, coordination, and communication can be supported by product Modeling application of tools for instance 3D modelling, 4D modelling, Building information modelling (Adriaanse et al., 2010). As web-based collaboration tools enhance the overall communication, according to Hosseini et al. (2017) team effectiveness will be improved as well because team effectiveness highly dependent on the quality of communication, and the quality of the information and exchanged data. Furthermore, Team resulted in improved data privacy (Comiskey et al., 2017). Consequently, improved information flow, elimination of various kinds of waste within the construction projects (Charalambous et al., 2017), and real-time visibility of project life cycle (Capterra, 2020) resulted in Improved collaboration (Ma et al., 2018).

Enhanced Workflow Management
Technical advancement in cloud computing and the web have fast-tracked rapid growth of globally dispersed project teams on construction projects . Monitoring and recording the progress of tasks, managing the flow of documents and information can be done by Workflow management application (Adriaanse et al., 2010) of web-based tools. Consistency level, efficiency, coordination, and quality improved due to workflows, best practice processes, document standards and metadata engaged by Common Data Environment (CDE) (Comiskey et al., 2017). BIM and digital technologies foster the integration of activities and strengthen the management of projects Data accuracy and better information management (East et al., 2008). In brief, web-based collaboration tools enhance collaboration through better management of workflow.

BIM Integration
The recent incorporation of integrated BIM for example online 3D model viewers enhanced communication through shared BIM which offers to cross-check more effectively (Charalambous et al., 2017). Collaboration tools integrating BIM can easily address Data sharing, access, and processing requirements issue of BIM adoption (Charalambous et al., 2017). BIM and related digital technology work as a catalyst to ensure more transparency, tighter integration, and increased productivity (Merschbrock & Munkvold, 2015).
The number of web-based collaboration tools in the present market is ranging from 200-375 according to different software reviewing online platforms (Capterra, 2018;Crowd, 2018;FinancesOnline, 2018b). This trend illustrates that over time, the number of tools is increasing, and features of tools are dynamically shifting to accommodate new user requirements, and technological advancement. As a consequence, new systems released by software vendors amplified the situation where uses have several different systems to fulfil their information requirements .

Previous studies and gap
Despite the extensive availability of collaboration tools, the problems regarding collaboration in construction remains an ongoing challenge (Mignone et al., 2016). Table 1 reports upon previous studies on collaboration technologies adopted in construction-related research to identify the focus and summary of findings.  Four key themes such as teamwork, building information modeling framework, evidence-based design practice, and modality have been identified as support of collaborative design.
Identified lack of a definitive framework of design collaboration.  (Abanda et al., 2015), collaboration tools validation (East et al., 2008) and interoperability issues (Zhang et al., 2017) have been explored. According to Oraee et al. (2017) collaboration in construction has been investigated mostly through the technology-oriented lens. Another group of researchers emphasized on the interrelation of collaboration tools and project teams and networks ( The aforementioned studies either considered technological aspects, advancements, working processes, adoption, implementation, barriers, success, and failure of collaboration tools. Most of the studies are either based on individual case studies or informal evidence provided by successful stories reported in the trade press. Moreover, studies listed in Table 1 either adopted a qualitative or exploratory approach. The former studies have identified the barriers and challenges of collaboration tools, and different technological aspects but the collaboration tools users' perceptions have hitherto not been considered in detail. As project-teams face various individual, environmental, and technological challenges while working with new technology or technology-based working processes (Merschbrock & Munkvold, 2015), the importance of exploring user perceptions of tools is vital. By identifying the knowledge gap, this study focused on investigating the user perceptions of collaboration tools adopters by conducting a quantitative text mining and qualitative content analysis approach.

RESEARCH METHODS
The study objectives necessitate exploring the end-users' perceptions of tools by direct quantitative analysis of reviews to identify patterns and latent connections of the different attributes. The research relies on analysis data related to the most common collaboration tools. The natural human language of the reviews is unstructured and required a method that will process the dataset to reveal patterns. Text mining can handle a large number of unstructured texts to reveal underlying patterns and trends. Furthermore, qualitative content analysis is conducted to investigate the dataset. This study explores the insights of users' reviews to unveil the underlying sentiments. A robust quantitative and qualitative analysis employing text mining and content analysis has been conducted. The positive and negative sentiments reviews were analysed through text mining to identify the most frequent words. Furthermore, Content analysis assisted to unearth the issues faced by the users.
Text mining is a remedial solution to discover knowledge from collections of unstructured text ). An increasing number of online reviews are posted daily on the internet which is a great source of data for making a variety of management decisions (Bi et al., 2019).

Data Collection
Five million, three hundred thousand users have been identified to use at least one of the collaboration tools (Capterra, 2018). This indicates the importance of collaboration tools but also the extent of usage. According to pertinent websites, ranking for the top ten tools is based either on customer number, social presence, price, ratings, or internally developed ranking algorithm (Capterra, 2018;Crowd, 2018;FinancesOnline, 2018a) Web-based construction collaboration tool listing websites like Capterra focused on a sponsor, highest-rated, and most reviews to rank the tools (Capterra, 2018). On the other hand of Financesonline ranking system is based on an internally developed SmartScore™ algorithm which has considered main functionalities, collaboration features, customization capabilities, available integrations, and so on for ranking the tools. Based on the ranking of Financesonline [viewed on 20.11.2018], the top six tools were CoConstruct, PlanGrid, Autodesk BIM 360, Procore, e-builder, and Aconex. The ranking contained within the website considered the collaboration features of the tools. For this reason, the aforementioned six tools were considered for analysing users' reviews. For extracting the reviews another review website 'Business Software Reviews from Software Advice™', was considered and accessed on 15.12.2018 to collect data of reviewers as this website accumulated reviews of a particular tool from a variety of sources (Softwareadvice, 2018).

Analyses
The study objectives necessitate exploring the end-users' perceptions of tools by direct quantitative analysis of reviews to identify patterns and latent connections of different attributes. The research relies on analysis data related to the most common collaboration tools using the same procedure adopted by  and (Miner et al., 2012). Details of the research techniques, design, and procedure are illustrated in Figure 1.

FIG 1: Research Methodology.
The natural human language of the reviews is unstructured and required a method that will process the dataset to reveal patterns. Text mining can handle a huge number of unstructured text and can discover knowledge . Furthermore, qualitative content analysis is conducted to investigate the dataset and the nature of data.

Information Retrieval
To create the data set, a web crawling method was implemented. The review website 'Business Software Reviews from Software Advice™', was accessed on 15.12.2018 to collect data of reviewers. Around 200 reviews for each tool resulted in 1,152 data set of reviews for the most popular six tools. Outwit hub (which is an open-data web scrapping tool) was implemented to extract data. Source code such as 'review data, 'company size' was utilized to collect the multiple web pages review data information. The data set was further filtered merged and organized in a spreadsheet. The data were organized in a spreadsheet containing thee columns: "collaboration tools name", "Company Size", and "Reviews". A total of 1152 Reviews of six prominent tools were prepared for further analysis.

Text Mining: Sentiment Analysis
Broad speaking, sentiment analysis is a set of techniques and tools aimed at detecting, extracting, and discovering the opinions and attitudes of authors of a text about certain entities (Mäntylä et al., 2018). These snippets of text typically reflect the feedback and reviews provided by users of the entity and are seen as 'a gold mine' of information. That is, these fragments of textual narrative in the form of reviews and feedbacks comprise subjective sentences that contain factual information, and reflect beliefs and views of users about an entity (Feldman, 2013). The most common application of sentiment analysis technique is for reviewing the experience of products and services users have written on websites devoted to discussions about the products or services at hand (Mäntylä et al., 2018).
Unstructured data can be processed using one of the available operators for sentiment analysis methods in RapidMiner like Rosette (Arianto et al., 2017). Having textual data as input, Rosette can return sentiment categories associated with an entire document, or for individual passages within a larger body of the text. Rosette relies on natural language processing (NLP) techniques for automated recognition and understanding of the view and opinions expressed in a human-generated text. It associates the subjective opinion embedded in a given text with a label: positive, negative, or neutral (Rosette, 2019). Prepared 1152 reviews were analysed in RapidMiner Studio 7.5 utilizing Rosette extension.

Text Mining: Term Frequency
Text mining is a process that discovers interesting and non-trivial knowledge from text documents (Ertek et al., 2014). Text mining is defined as "an attempt to separate valuable keywords from a mass of other words" to identify meaningful patterns . For this present study, text mining analysis was conducted in RapidMiner Studio 7.5, which is an open-source data mining and business analytics software solution. After conducting the sentiment analysis, the positive and negative datasets were further analysed and processed to identify the term frequency.

Content Analysis
The qualitative content analysis determines the specific word frequency appear in a text and can assist to describe the meaning of the textual narrative. Content analysis has been defined as "a research technique for making replicable and valid inferences from text to the context of their use" (Yu et al., 2006). In case of the collaboration tools users' perception, after conducting quantitative analysis through the text mining approach, content analysis was adopted to further investigate the most frequently appearing words connection and association within the positive and negative data set. According to Fellows and Liu (2015), content analysis is an appropriate data analysis technique in management and construction research.

Sentiment Analysis
Among the 1,152 data sample frame, 379 of cases represent large companies, whereas 353 belong to small-sized companies and 205 are representing medium-sized companies. Only 16 of the reviewers belong to micro companies as illustrated in Figure 2. The remaining sample did not have any information. This sample represents a reasonable balance of participants and coverage of all categories of major users of collaboration tools in the construction industry, given that due to resource limitations, micro companies are not among typical users of collaboration tools. Among the reviews, 804 were positive about the collaboration tools, only 26 cases were neutral and 322 were negative as indicated in Figure 3.

Text Mining (Term Frequency)
Term Document Matrix (TDM) is a procedure that converts textual data into a TDM . Words are operated into "tokens" as text mining algorithms treat words in a sentence as unrelated objects . According to , the tokenizing process is to convert text into bags of tokens and tokens create the TDM, in which each token is an attribute and each document is a case. Figure 4 indicated the steps of TDM creation followed by Filtering stopwords which removed common terms for instance "a", "and", "etc." Large complex data sets of textual documents contain a substantial amount of irrelevant and noisy information (Ertek et al., 2014). A frequency-based feature application has been utilized to create meaningful tokens and to remove noise for this particular dataset. This resulted in a Document containing Total occurrence and Document occurrence of most frequent words both in positive and negative data set.

Content Analysis of Positive Dataset
Succeeding the sentiment analysis, the frequency of words that appeared most were thematically grouped into a positive and negative data set (refer to Tables 2 and 3 respectively). For the large positive data set, a list of 90 words that have frequently appeared have been identified. Among the list word like 'great', 'construct', have not considered. Among 90 words identified, those words that have appeared more than 100 times are provided in Table  2, along with total number of occurrences and document occurrences. Topmost frequently mentioned six words have been considered for content analysis. Project management tool, construction project management tool, project delivery, integrated project, project management coordination, project need, project team, project coordinator, clicking between projects, project-related documents, project organization, Accessibility of project information, project stakeholders and total control over projects.
"worked on a number of large complex projects using the tool and it was very useful, all drawings in one place and you can see previous versions of the same drawing and how it has progressed." According to Case 12 "A must-have software for medium to large scale projects best for traceability and sharing of large files." The above example indicated that users expressed their positive perceptions of the tools based on how they helped them to manage different types and sizes of tools along with the different features, aspects, levels and management issues of the projects. Communication aspect of tools have been mentioned also viz Case 65: "The software allows Project Management teams to efficiently and effectively communicate project details or issues, collaborate on solutions to issues, and have access to any project details at any time there is an internet connection". Moreover, project financial tracking, accessibility of project information, project stakeholders, and ease of managing a project also mentioned to indicate positive sentiment.
Great: Great term total occurrence is 414 and document occurrence is 255. This term mainly associated to indicate how positivity of the reviewers regarding tools. Great also used to highlight some features of tools that users liked most such as user-friendliness, collaboration, streamline communication, workflow, etc features indicated as great.
For instance, case 45 mentioned "It's a really great program and I would not be able to work remotely without it. You can access the job details anywhere. You can update items on the go. And it is relatively easy to use." On the other hand case, 99 indicated "….great for keeping track of communications, tasks, RFIs, etc., maintaining a directory, storing documents and drawings". Moreover, case 102, mentioned the positive aspect of tools as a "great support network", and case 233 as a "Great application to impress your clients….and Great for organizing tight schedules by the weeks." Software: This term's total occurrence is 409 and document occurrence is 255. Associated words included 'estimation', 'learning' and 'integration'. Case 7 indicated that: "It is hard to break the traditional methods (i.e., email and project folders on the hard drive), especially for those that don't like learning new software,'' and in so doing emphasized the learning aspect of tools. Case 8 explained the features of tool that they like viz: "estimate in the software, which then translates to a completed Specs/Selections sheet for the customers." In brief the software word association indicated the tools features, users' perception of tools and different integration of tools. Such as Case 112 i.e. "software useful to track and documents all kinds of items to ensure proper work is done." Document: This term's total occurrence is 333 and document occurrence is 182. The word 'document' word is associated with terms such as 'contract documents', 'reference documents', 'drawing documents', 'document management', 'document control' and 'document traceability'. The word document is frequently mentioned as users talked about the tools document management features, how the tools deal with different sort of documents and their experiences of documents and tools. For example, Case 90 indicated the positive aspect of tools by mentioning the: "Best tool for the job site for having contract documents handy", case 21 indicated the searchability and document transmission features is the positive aspect of tools. Another case 23 indicated the "fully complies with internal needs and also allows the organization to be up-to-date with ISO 30300 and the best practices in the documentary field." Tracking features of documents and monitoring of emails is vital. Document register, request for information (RFI), was mentioned by case 805. In brief, document management, document control, document tracking, document associated tasks are vital for the users and the tools effectiveness is judged by its overall document management aspect.
Manage: This term's total occurrence is 269 and document occurrence is 193. Content analysis reveal that the word 'manage' words is used in association with terms such as 'cloud-based management, 'a user-friendly file management system', 'better time management', 'better drawing control management', 'documentary management' and 'workflow and correspondence management'. According to the way the 'manage' word frequently appeared it indicated that users' perceptions of tools capability to manage the project aspect and the overall management issues are vital. For example, Case 120 mentioned that "Client Management via the tools removes the 'grey area' often used by clients to negotiate free services." Another Case 4 indicated the file management system to explain the positive aspect of tools such as "It is a very user-friendly file management system that helps us/our Client keep track of all the pertinent file and drawing document on a project by project (task by task)." Conversely, Case 766 indicated "being able to manage a project on one platform that is web-based allowing for remote access", incorporated the accessibility of tools and how they manage projects.

Construct:
The total occurrence of construct term is 251 and document occurrence is 170. Construct term often mentioned to indicate construction collaboration tools, construction management software, construction type, construction project, and construction process and construction document, and so on. For instance, case 39 mentioned positive aspects of tools as "Best construction project management tool, ease of use, customer support, all field tools, financial tools, project management tools, closeout tools etc." Another case 68 said about tools "…helping me to organize my construction project by providing subcontractors a platform for consolidated document review." In short, construct terms often used to indicate projects, tools, processes and so on.
Custom: Total occurrence of 'custom' is 223 and document occurrence is 158. The word custom is further associated with such as 'customer service', 'customization' and 'customize'. Users frequently utilize this word to discuss the customer service facilities provided by the tools followed by the customization capabilities of different tools. Examples include Case 66 who indicated that: "the customer service by far is the BEST I've ever encountered anywhere. They ALWAYS get you an answer in such a timely manner and are so polite, friendly and willing to go the extra mile." Case 422 indicated that: "so the Customer support is a big help to assist the users of the functionality, custom builder, flexibility in customization, connect with customers." Another reviewer mentioned: "The ability to bring everything together into one file from communication with trades/suppliers/Customers to filing contracts/change orders/financial matters. Business Intelligence (BI) reporting provides more options for customized reporting." So, customization features are vital for users and not all tools provide the features.

Feature:
The 'feature' word total occurrence is 216 and document occurrence is 166 and was associated with features such as 'workflow features'. According to case 213 "The snapshot feature allows me to send changes to foremen in the field when I receive it." Other features that users discussed included: "The features I like most about this application are the ability to sync to other users.", "tracking features", "Time tracking features", "good features in computing estimation" and "document management features." Moreover, "The punch list feature" is incredibly convenient according to Case 321. Another Case mentioned that "constantly growing and adding new features, multiple functions" was desirable. These indicated the features of collaboration tools are vital for the users and the overarching positive aspect of tools were their features that satisfied the users' needs.

Content Analysis of Negative Dataset
After sentiment analysis, the negative and positive data set was investigated by text mining approach and the frequency of total word and document occurrence of word have been created. In the negative data set list word like 'great', 'builder' and 'construct' were omitted as they do not exhibit any meaningful result. Although a total of 81 frequently appearing words were noted, only words had more than 40 total occurrences and were selected for content analysis. Findings of the top six frequently appearing words are now discussed in detail. Project management, large-size projects, project documents, project manager, manage an entire project and tracking project.

Process:
The total occurrence of process term is 68 and document occurrence is 40. Negative sentiment reviews revealed associated concept with "process " terms are "Slow the process down", "process need work", "streamlining process", "selection process", "construction process", "process integration", "billing process", "business process", "can be slow to response and process", "not a smooth process", and so on.
Associated concepts of process term within negative data set indicated various process-related issues which resulted in difficulties. For example, case 150 indicated "At times, the workflows can slow the process down. You end up waiting on other people to complete their tasks". Furthermore, case 314 mentioned, "Selection process is painstaking; No import option for price catalogs or bidding software". Overall, the frequency of the process term indicates process-related negative sentiment of users.

DISCUSSION AND CONCLUSION
The overall sentiment analysis revealed that users incur both positive and negative experiences of collaboration tools. This is an intuitive finding, given the broad range of knowledge and experience within the target population frame. Although existing tools exist in large number, still users are face problems. Overall, the findings indicated that the document management features of tools are significant, and users face issues regarding document findings, searches, and organization and file format aspects. For instance, "Document" word in Table 1 (Total occurrence 333) indicated that tools are judged by their overall document management aspect. However, in Table 2 the same word "Document" is associated with negative sentiment of tools indicating a variety of difficulties related to the document management features. Report features were also observed to be important to users. Issues and problems related to report features were identified and several users discussed the customization of report features. On the other hand, based on "project size" collaboration tools users experience, and tools adoption differs. Apart from these, customer service assistance of tools is an important criterion for users' satisfaction. Moreover, userfriendliness of tools is also shown to be of significance.
The study investigated the users' perception of collaboration technology and contributes to the wider body of knowledge in this area. By gaining a deeper and richer appreciation of both the negative and positive perceptions, industry practitioners are better able to implement strategies to maximize collaboration potential in their businesses. In turn, such knowledge could facilitate improvements in productivity performance on site and facilitate better business outcomes. The results present an insight on users' perceptions regarding collaboration tools. The methodology adopted investigated big data and created a base for further analyzing the dataset. This methodology studied the open source domain data and thus analysis is reproducible. The study also identified the points against which collaboration tools proved satisfactory or unsatisfactory users. The study has limitations, in that it considered only the most frequent terms. Moreover, consideration of other groups of words and different tools may result in broader findings. Thus, wider sample may be considered in a future study to investigate users' perception in greater detail.