Additional detail on prior versions #20 (#21)

* Additional detail on prior versions #20

* add opensource alternative to resolve #7

Author: muctadir

paper/paper.md

@@ -76,15 +76,15 @@ Qualitative analysis of data is relevant for a variety of domains including empi
 
 Analyzing qualitative data has been proven to be labor intensive and time consuming task [@analysing_qual_data] due to its nature. Thematic analysis [@thematic_analysis] is a powerful yet flexible method for performing such analysis. Analyzing textual data through this method allows a researcher to understand experiences and thoughts, as well as emotions and behaviors throughout a data set. Due to the flexibility of this analysis method, the users are not bound to using only one paradigmatic perspective but within different data sets can use different ones [@clarke_psych].
 
-As thematic analysis is a widely used qualitative analysis technique, several commercial tools are available, such as Atlas.ti^[https://atlasti.com/] and maxQDA^[https://www.maxqda.com/]. Although these tools are very well developed, there are three major trade offs that inspired the development of LaMa.
+As thematic analysis is a widely used qualitative analysis technique, several commercial tools are available, such as Atlas.ti^[https://atlasti.com/] and maxQDA^[https://www.maxqda.com/]. We also investigated open-source applications that allows labelling of artifacts such as Label Studio^[https://labelstud.io/]. Although these tools are very well developed, there are three major trade offs that inspired the development of LaMa.
 
 - **Cost:** As these are commercial tools, their services are not free and can be quite expensive depending on the subscription.
 
 - **Data access and privacy:** Qualitative researches often process sensitive data, such as legally protected information, private information of individuals. With rising privacy concerns, increasing number of research organizations are requiring specialized approval for working with such data. For example, at Eindhoven University of Technology it is mandatory, among other information, to specify which individuals can have access to the research data. With commercial tools, control over the access of the research data or the storage location are often unavailable.
 
 - **Complex collaboration workflow:** Collaborative labelling or coding is an established method for reducing bias during qualitative analysis [@APracticalGuidetoCollaborativeQualitativeDataAnalysis]. While commercial tools provide this feature in various forms, the process for resolving conflicting labels is often complicated.
 
-Based on these points we developed LaMa, which is a web application intended to support the thematic analysis and is built based on an existing application called the Labeling Machine [@labeling_machine], which is forked from the an earlier Labeling Machine [@labeling_machine_orig]. Its key features are described in the following section.
+Based on these points we developed LaMa, which is a web application intended to support the thematic analysis and is built based on an existing application called the Labeling Machine [@labeling_machine], which is forked from the an earlier Labeling Machine [@labeling_machine_orig]. In addition to significantly improving the user interface, LaMa provides additional features such as multi-labelling, hierarchical theming, and change tracking. Its key features are described in the following section.
 
 # Key features
 
@@ -96,14 +96,14 @@ LaMa in its core functionality is similar to comparable labelling tools for qual
 
 - **Collaborative labelling:** With LaMa multiple researchers can simultaneously label same set of text artifacts. During labelling newly created labels are immediately shared with other labelers, which facilitates the reuse of existing labels. Furthermore, a LaMa project can be configured so that it requires one artifact to be labeled by more than one labeler to reduce individual bias.
 
+- **Multi labelling:** LaMa allows each artifact to be labelled with multiple labels. This feature is particularly useful if a researcher wants to label an artifact from more then one viewpoints. These viewpoints, which are called label types, can be configured during project creation.
+
 - **Conflict resolution:** To ensure consensus during collaboration, automatic conflict detection has been implemented. A conflict occurs when one artifact is labeled differently by multiple labelers. LaMa users can view these disagreements, which facilitates a dialog among corresponding labelers to agree on a label. Having this dialog and resolving the disagreement are very important for reducing individual bias during the labelling process.
 
 - **Themes:** LaMa allows users to group labels into themes, thereby providing help in the classification and the analysis of the data. Furthermore, themes can be categorized hierarchically further aiding in the analysis and classification process.
 
 - **Traceability:** LaMa keeps a record of all the changes made to the artifacts, labels and themes. These changes are visible on the details page of the corresponding entities. This adds an extra layer of traceability.
 
-- **Visualizing hierarchy:** To obtain insight into how artifacts are mapped to labels, and labels ordered into themes, a visualization of the hierarchy can be constructed on demand.
-
 # Conclusion and future work
 
 LaMa is an open-source web-application for thematic labelling of qualitative data. Its key-assets are obtaining insight into data by hierarchically grouping them into labels & themes and facilitating better collaboration between users through features such as collaborative labelling & conflict resolution.