fin

text/0_abstract.tex

@@ -1,7 +1,7 @@
 \chapter*{Abstract}
 \label{cha:abstract}
 
-StackExchange is a question and answer platform and as many other social platforms, StackExchange is eager to provide a good first impression to users. StackExchange made many decisions to attract new users. One of these decisions was to introduce the \emph{new contributor} indicator which is shown to users that may answer a question from a new user. This thesis investigates whether this change improved the impression, new users experience. To measure whether the change achieved its intended target, this thesis uses VADER to quantify the sentiment of the answers to questions of new contributors which are then used in an interrupted time series analysis. The results indicate that in some of the communities the change did indeed achieve its intended purpose.
+StackExchange is a question and answer platform and as many other social platforms, StackExchange is eager to provide a good first impression to users. StackExchange made many decisions to attract new users. One of these decisions was to introduce the \emph{new contributor} indicator which is shown to users that may answer a question from a new user. This thesis investigates whether this change improved the experience of new users. To measure whether the change achieved its intended target, this thesis uses VADER to quantify the sentiment of the answers to questions of new contributors which are then used in an interrupted time series analysis. The results indicate that in some of the communities the change did indeed achieve its intended purpose.
 
 
 %This is a place-holder for the abstract. 

text/2_relwork.tex

@@ -8,7 +8,7 @@ StackExchange\footnote{\url{https://stackexchange.com}} is a community question
 
 Originally, StackExchange started with StackOverflow\footnote{\url{https://stackoverflow.com}} in 2008\footnote{\label{atwood2008stack}\url{https://stackoverflow.blog/2008/08/01/stack-overflow-private-beta-begins/}}. Since then StackExchange grew into a platform hosting sites for 174 different topics\footnote{\label{stackexchangetour}\url{https://stackexchange.com/tour}}, for instance, programming (StackOverflow), maths (MathOverflow\footnote{\url{https://mathoverflow.net}} and Math StackExchange\footnote{\url{https://math.stackexchange.com}}), and typesetting (TeX/LaTeX\footnote{\url{https://tex.stackexchange.com}}). Questions on StackExchange are stated in the natural English language and consist of a title, a body containing a detailed description of the problem or information needed, and tags to categorize the question. After a question is posted the community can submit answers to the question. The author of the question can then accept an appropriate answer which satisfies their question. The accepted answer is then marked as such with a green checkmark and shown on top of all the other answers. Figure \ref{soexamplepost} shows an example of a StackOverflow question. Questions and answers can be up-/downvoted by every user registered on the site. Votes typically reflect the quality and importance of the respective question or answers. Answers with a high voting score raise to the top of the answer list as answers are sorted by the vote score in descending order by default. Voting also influences a user's reputation \cite{movshovitz2013analysis}\footref{stackexchangetour}. When a post (question or answer) is voted upon, the reputation of the poster changes accordingly. Furthermore, downvoting of answers also decreases the reputation of the user who voted\footnote{\url{https://stackoverflow.com/help/privileges/vote-down}}.
 
-Reputation on StackExchange indicates how trustworthy a user is. To gain a high reputation value a user has to invest a lot of time and effort to reach a high reputation value by asking good questions and posting good answers to questions. Reputation also unlocks privileges which may differ slightly from one community to another\footnote{\url{https://mathoverflow.com/help/privileges/}}\mfs\footnote{\url{https://stackoverflow.com/help/privileges/}}. 
+Reputation on StackExchange indicates how trustworthy a user is. To gain a high reputation value a user has to invest a lot of time and effort to reach by asking good questions and posting good answers to questions. Reputation also unlocks privileges which may differ slightly from one community to another\footnote{\url{https://mathoverflow.com/help/privileges/}}\mfs\footnote{\url{https://stackoverflow.com/help/privileges/}}. 
 With privileges, users can, for instance, create new tags if the need for a new tag arises, cast votes on closing or reopening questions if the question is off-topic or a duplicate of another question, or when a question had been closed for no or a wrong reason, or even get access to moderation tools. 
 StackExchange also employs a badge system to steer the community\footnote{\label{stackoverflowbadges}\url{https://stackoverflow.com/help/badges/}}. Some badges can be obtained by performing one-time actions, for instance, reading the tour page which contains necessary details for newly registered users, or by performing certain actions multiple times, for instance, editing and answering the same question within 12 hours.
 Furthermore, users can comment on every question and answer. Comments could be used for further clarifying an answer or a short discussion on a question or answer.
@@ -355,7 +355,7 @@ Linguistic Inquiry and Word Count (LIWC) \cite{pennebaker2001linguistic,pennebak
 % - very old (1966), continuously refined, still in use (vader)
 % - misses lexical feature detection (acronyms, ...) and sentiment intensity (vader)
 
-General Inquirer (GI)\cite{stone1966general} is one of the oldest sentiment tools still in use. It was originally designed in 1966 and has been continuously refined and now consists of about 11000 words where 1900 positively rated words and 2300 negatively rated words. Like LIWC, GI uses a polarity-based lexicon and therefore is not able to capture sentiment intensity\cite{hutto2014vader}. Also, GI does not recognize lexical features, such as acronyms, initialisms, etc.
+General Inquirer (GI)\cite{stone1966general} is one of the oldest sentiment tools still in use. It was originally designed in 1966 and has been continuously refined. Now it consists of about 11000 words with 1900 positively rated words and 2300 negatively rated words. Like LIWC, GI uses a polarity-based lexicon and therefore is not able to capture sentiment intensity\cite{hutto2014vader}. Also, GI does not recognize lexical features, such as acronyms, initialisms, etc.
 
 
 %Hu-Liu04 \cite{hu2004mining,liu2005opinion}, 2004
@@ -425,7 +425,7 @@ Word-Sense Disambiguation (WSD)\cite{akkaya2009subjectivity} is not a sentiment
 %updateing (extend/modify) hard (e.g. new domain) (vader)
 
 \textbf{Machine Learning Approches}\\
-Because handcrafting sentiment analysis requires a lot of effort, researchers turned to approaches that offload the labor-intensive part to machine learning (ML). However, this results in a new challenge, namely: gathering a \emph good data set to feed the machine learning algorithms for training. Firstly, \emph good data set needs to represent as many features as possible, otherwise, the algorithm will not recognize it. Secondly, the data set has to be unbiased and representative of all the data of which the data set is a part of. The data set has to represent each feature in an appropriate amount, otherwise, the algorithms may discriminate a feature in favor of other more represented features. These requirements are hard to fulfill and often they are not\cite{hutto2014vader}. After a data set is acquired, a model has to be learned by the ML algorithm, which is, depending on the complexity of the algorithm, a very computationally-intensive and memory-intensive process. After training is completed, the algorithm can predict sentiment values for new pieces of text, that it has never seen before. However, due to the nature of this approach, the results cannot be comprehended by humans easily if at all. ML approaches also suffer from a generalization problem and therefore cannot be transferred to other domains without accepting a bad performance, or updating the training data set to fit the new domain. Updating (extending or modifying) the model also requires complete retraining from scratch. These drawbacks make ML algorithms useful only in narrow situations where changes are not required and the training data is static and unbiased.
+Because handcrafting sentiment analysis requires a lot of effort, researchers turned to approaches that offload the labor-intensive part to machine learning (ML). However, this results in a new challenge, namely: gathering a \emph good data set to feed the machine learning algorithms for training. Firstly, a \emph {good data set} needs to represent as many features as possible, otherwise, the algorithm will not recognize it. Secondly, the data set has to be unbiased and representative of all the data of which the data set is a part of. The data set has to represent each feature in an appropriate amount, otherwise, the algorithms may discriminate a feature in favor of other more represented features. These requirements are hard to fulfill and often they are not\cite{hutto2014vader}. After a data set is acquired, a model has to be learned by the ML algorithm, which is, depending on the complexity of the algorithm, a very computationally-intensive and memory-intensive process. After training is completed, the algorithm can predict sentiment values for new pieces of text, that it has never seen before. However, due to the nature of this approach, the results cannot be comprehended by humans easily if at all. ML approaches also suffer from a generalization problem and therefore cannot be transferred to other domains without accepting a bad performance, or updating the training data set to fit the new domain. Updating (extending or modifying) the model also requires complete retraining from scratch. These drawbacks make ML algorithms useful only in narrow situations where changes are not required and the training data is static and unbiased.
 
 % naive bayes
 % - simple (vader)
@@ -445,7 +445,7 @@ Support Vector Machines (SVM) use a different approach. SVMs put data points in
 
 %generall blyabla, transition to vader
 
-In general, ML approaches do not provide an improvement over hand-crafted lexicon approaches as they only shift the time-intensive process to training data set collections. Furthermore, lexicon-based approaches seem to have progressed further in terms of coverage and feature weighting. However, many tools are not specifically tailored to social media text analysis and leak in coverage of feature detection. 
+In general, ML approaches do not provide an improvement over hand-crafted lexicon approaches as they only shift the time-intensive process to training data set collections. Furthermore, lexicon-based approaches seem to have progressed further in terms of coverage and feature weighting. However, many tools are not specifically tailored to social media text analysis and lack in coverage of feature detection. 
 
 %vader (Valence Aware Dictionary for sEntiment Reasoning)(grob) \cite{hutto2014vader}
 % - 2014

text/3_method.tex

@@ -21,7 +21,7 @@ This thesis investigates the following criteria to determine whether the change
  \item \textbf{Vote score of questions}. This symbolizes the feedback the community gives to a question. Voters will likely vote more positively (not voting instead of down-voting, or upvoting instead of not voting) due to the \emph{new contributor} indicator. Thereby the vote score should increase after the change.
  \item \textbf{Amount of first and follow-up questions}. This symbolizes the willingness of users to participate in the community. Higher amounts of first questions indicate a higher number of new participating users. Higher follow-up questions indicate that users are more willing to stay within the community and continue their active participation.
 \end{itemize}
-If these criteria improve after the change is introduced, the community is affected positively. If they worsen, the community is affected negatively. If the criteria stay largely the same, then the community is unaffected. Here it is important to note that a question may receive answers and votes after the \emph{new contributor} indicator is no longer shown and therefore these are not considered part of the data set to analyze.
+If these criteria improve after the change is introduced, the community is affected positively. If they worsen, the community is affected negatively. If the criteria stay largely the same, then the community is unaffected. Here it is important to note that a question may receive answers and votes after the \emph{new contributor} indicator is no longer shown. Therefore these answers and votes are not considered part of the data set to analyze.
 %only when new contributor insicator is shown
 
 

text/5_results.tex

@@ -2,7 +2,7 @@
 
 %TODO inconsistent tenses
 
-This section shows the results of the experiments described in section 3 on the data sets described in section 4. In the following pages, there 3 diagrams for each community. The diagrams capture 3 different aspects: the sentiment of answers, the vote score of questions, and the number of questions. These aspects are all measured concerning questions from new users.
+This section shows the results of the experiments described in section 3 on the data sets described in section 4. In the following pages, there are 3 diagrams for each community. The diagrams capture 3 different aspects: the sentiment of answers, the vote score of questions, and the number of questions. These aspects are all measured concerning questions from new users.
 
 In diagrams (a), the blue line states the average sentiment (\emph{average sentiment} in diagram legend) of the answers to questions from new contributors. Also, the numbers attached to the blue line indicate the number of answers to questions from new users that formed the average sentiment. The orange line (\emph{sm single ITS} in the diagram legend) represents the ITS over the whole period of the available data. As stated in section 3.2, data density variability is a factor to take into account, therefore, the orange line represents the weighted ITS. The green, red, purple, and brown lines also represent weighted ITS, however, the time periods considered for ITS before and after the change are limited to 6, 9, 12, and 15 months respectively.
 
@@ -118,7 +118,7 @@ On stats.stackexchange.com the average sentiment decreases steadily prior to the
 
 The vote score also decreases prior to the change and decreases even faster afterward. However, 4 to 5 months after the change, the vote score falls into a valley for about 10 months before recovering. This can be the result of another outside factor. 
 
-By looking at the number of 1st questions, it can be said that the vote score dipped because the number of first questions spiked during the previously stated time frame. As the community suddenly receives over-proportionally many new users, the quality of interactions in the community drops for a short time. This theory would be supported by \cite{lin2017better}. While the trends for 1st and follow-up questions are stagnant before the change they improved after the change. This indicates an increase in contributions from new users and that the change works for this community.
+By looking at the number of 1st questions, it can be said that the vote score dipped because the number of first questions spiked during the previously stated time frame. As the community suddenly receives a disproportionally many new users, the quality of interactions in the community drops for a short time. This theory would be supported by \cite{lin2017better}. While the trends for 1st and follow-up questions are stagnant before the change they improved after the change. This indicates an increase in contributions from new users and that the change works for this community.
 
 In summary, the sentiment improves after the change, the vote score is inconclusive, and the number of 1st and follow-up questions improves, indicating the community benefits from the change.
 \begin{figure}[H]
@@ -149,9 +149,9 @@ In summary, the sentiment improves after the change, the vote score is inconclus
 \section{tex.stackexchange.com}
 On tex.stackexchange.com the average sentiment is low compared to the other investigated data sets. Prior to the change the average sentiment only slightly decreases. When the change is introduced the regression takes a dip down and after the change, the average sentiment increases drastically, indicating the change has a positive effect on the community. Future data will be required to see if this upward trend continues or evens out. 
 
-In stark contrast, the vote score shows a downward trend. The vote score is on a continuous downward trend with a peek around the change date but the vote score does not improve in the long term. Although there is a short window around the change date where vote scores are higher compared to before and after the change, this is not a result of the change but a coincidence. The vote score increases several months before the change actually occurs. The continuous downward trend with a peek around the change date does not indicate that the vote score improves in the long term. Either way, this indicates the change did not affect the vote score. 
+In stark contrast, the vote score shows a downward trend. The vote score is on a continuous downward trend with a peak around the change date but the vote score does not improve in the long term. Although there is a short window around the change date where vote scores are higher compared to before and after the change, this is not a result of the change but a coincidence. The vote score increases several months before the change actually occurs. The continuous downward trend with a peek around the change date does not indicate that the vote score improves in the long term. Either way, this indicates the change did not affect the vote score. 
 
-The amount of 1st questions improved after the change and turned the downward trend into an upward trend with the same grade. The number of follow-up questions does not see an improvement and continues the downward trend like before the change. This shows that more new contributors ask their 1st question than before, however, they still tend to become one-day-flies \cite{slag2015one}. Also, the number of the 1st questions, the months of -44, -32, -20, -8, 4, and 16 are local minima, indicating seasonality in the data \cite{bernal2017interrupted}. These months are all in December when the people of large parts of the world are on holiday.
+The amount of 1st questions improved after the change and turned the downward trend into an upward trend with the same grade. The number of follow-up questions does not see an improvement and continues the downward trend like before the change. This shows that more new contributors ask their 1st question than before, however, they still tend to become one-day-flies \cite{slag2015one}. Also, the months of -44, -32, -20, -8, 4, and 16 are local minima for the number of 1st questions, indicating seasonality in the data \cite{bernal2017interrupted}. These months are all in December when people of large parts of the world are on holiday.
 
 In summary, the sentiment improves, the vote score is unaffected, and the number of 1st questions does improve, suggesting that the community benefits from the change.
 \begin{figure}[H]
@@ -183,7 +183,7 @@ In summary, the sentiment improves, the vote score is unaffected, and the number
 \section{unix.stackexchange.com}
 On unix.stackexchange.com the average sentiment decreases prior to the change. When the change is introduced the regression takes a small dip down, however, the average sentiment increases fast after the change, indicating the change has a positive effect. 
 
-The vote score shows a continuous downward trend. At the change date, the trend does not even move by much and continues downward at about the rate, indicating the change does not affect the vote score in this community. 
+The vote score shows a continuous downward trend. At the change date, the trend does not move by much and continues downward at about the same rate, indicating the change does not affect the vote score in this community. 
 
 The number of 1st questions improved after the change and turned the stagnant trend into an increasing trend. The number of follow-up questions also improved in a similar manner. This shows that new contributors ask more questions than before. 
 
@@ -325,11 +325,11 @@ In summary, the sentiment does not seem to be affected. The vote score continues
 % more data in the future will be required to determine if upward trend in the end continues
 
 \section{SuperUser.com}
-SuperUser shows only sightly decreasing average sentiment and vote score up to the change. At the change time the regressions take a dip down and the regression shows a downward trend after the change. However, the huge drop in sentiment and vote score does not align with the change date but happens 4 months after the change. 
+SuperUser shows only sightly decreasing average sentiment and vote score up to the change. At the change time the regressions takes a dip down and the regression shows a downward trend after the change. However, the huge drop in sentiment and vote score does not align with the change date but happens 4 months after the change. 
 
 In the same time frame the number of 1st questions skyrockets to more than triple the previous levels. This is similar to the feature found in the results from stats.stackexchange.com, although this example is much more pronounced. This feature also seems to be produced by the huge influx of new users to the community. As described in \cite{lin2017better}, the quality of interactions in the community dip for a while but recovers over time. The sentiment recovers after about 13 months. The vote score also starts to recover at the same time, but not as quickly as the sentiment value. Due to this spike in the number of new users, the analysis does not yield any meaningful results.
 
-The number of 1st questions decreases prior to the change and then goes through the roof indicating a huge wave of new users indicating a drastic influx of new users. Data available in the future will show if the recovery at the end of the timeframe is persistent. Even though a lot of new users joined the community, the number of follow-up questions stayed largely the same. 
+The number of 1st questions decreases prior to the change and then increases dramatically, indicating a drastic influx of new users. Data available in the future will show if the recovery at the end of the timeframe is persistent. Even though a lot of new users joined the community, the number of follow-up questions stayed largely the same. 
 
 In summary, the sentiment and vote score analysis does not yield a meaningful result as the time frame after the change includes an outside factor with a huge impact. The number of follow-up questions does not seem to increase despite the number of first questions doubling, indicating that a lot of the new users are one-day-files\cite{slag2015one}. The results of this analysis are inconclusive.
 \begin{figure}[H]
@@ -371,4 +371,4 @@ When looking at the results of SuperUser, the community stands out and shows int
 
 
 \section*{Summary}
-In summary, the change introduced by StackExchange clearly improved the engagement in 6 of the 10 investigated communities. Sentiment, vote score, and number (1st and follow-up) questions rose as a result. The other 4 communities do not profit from the change. Although, many statistics jump up to a higher level the downward trends are not stopped. The statistics of SuperUser show a large influx of new users about 6 months after the change sending the sentiment and vote score on a deep dive and with the decrease in new users they raise again. However, this event is not related to the change and the magnitude of the change in new user numbers renders the analysis incomparable.
+In summary, the change introduced by StackExchange clearly improved the engagement in 6 of the 10 investigated communities. Sentiment, vote score, and the number of (1st and follow-up) questions rose as a result. The other 4 communities do not profit from the change. Although, many statistics jump up to a higher level the downward trends are not stopped. The statistics of SuperUser show a large influx of new users about 6 months after the change sending the sentiment and vote score on a deep dive and with the decrease in new users they raise again. However, this event is not related to the change and the magnitude of the change in new user numbers renders the analysis incomparable.

text/6_discussion.tex

@@ -24,7 +24,7 @@ The other 4 communities do not seem to profit from the change so clearly: Mathov
 
 The average sentiment stays constant on MathOverflow before the change and decreases afterward. The sentiment levels start increasing six months before the change and are unrelated. However, the sentiment falls sharply at the change date, indicating the sentiment values are affected negatively by the change. The vote score is steadily increasing before the change and crashes shortly after the change. However, the vote score is very high compared to other communities. The number of 1st questions stabilizes after the change compared to the slight downward previously. By looking at the graph of the 1st questions, the months of -41, -29, -17, -5, and 7 are local maxima, indicating seasonality in the data \cite{bernal2017interrupted}. These months are all in March. Also while the number of 1st questions stabilizes to a constant trend, the number of follow-up questions decreases, indicating that the new users tend more to become one-day-flies as time passed on \cite{slag2015one}.
 
-math.stackexchange.com shows a downward trend before and after the change in sentiment and vote score. The sentiment ITS is particularly affected by the low sentiment values at the end and future data is required to determine if this trend continues. However, the number of 1st questions stabilizes a bit after changes, and follow-up questions even see a slight increase after the change. The graph with the number of questions from new contributors shows a good example of seasonality in data \cite{bernal2017interrupted}. The month 0 indicates August. For the 1st questions, the months -44, -32, -20, -8, 4, and 16 are all a local minimum. These months are all in December. Similarly, the months -38 and -37, -26 and -25, -14 and -13, -2 and -1, and 10 and 11 are all in June and July. During both these times, the people large portions of the world are going through a holiday season which may likely explain these regular dips in contribution. The graph for the follow-up questions also shows dips at the same times, although the dips in December are not always as discernible.
+math.stackexchange.com shows a downward trend before and after the change in sentiment and vote score. The sentiment ITS is particularly affected by the low sentiment values at the end and future data is required to determine if this trend continues. However, the number of 1st questions stabilizes a bit after the change, and follow-up questions even see a slight increase after the change. The graph with the number of questions from new contributors shows a good example of seasonality in data \cite{bernal2017interrupted}. The month 0 indicates August. For the 1st questions, the months -44, -32, -20, -8, 4, and 16 are all a local minimum. These months are all in December. Similarly, the months -38 and -37, -26 and -25, -14 and -13, -2 and -1, and 10 and 11 are all in June and July. During both these times, the people large portions of the world are going through a holiday season which may likely explain these regular dips in contribution. The graph for the follow-up questions also shows dips at the same times, although the dips in December are not always as discernible.
 
 The electronics.stackexchange.com community has a similar pattern for the sentiment value and vote scores compared to math.stackexchange.com. However, the sentiment values seem to recover after about 12 months and future data is required to see if the rise at the end of the period is a long-term trend. The rising number of first questions of new contributors stops at the change date and transitions into a decreasing pattern. However, the number of follow-up questions increases slightly after the change. Even though the number of new users decreases after the change the amount of follow-up questions increases, indicating the number of one-day-flies decreases \cite{slag2015one}.
 

text/main.tex

@@ -160,8 +160,8 @@
 \newcommand{\myhomestreet}{Petergasse~23/13} %% your home street (with house number)
 \newcommand{\myhometown}{Graz} %% your home town
 \newcommand{\myhomepostalnumber}{8010} %% your postal number of home town
-\newcommand{\mysubmissionmonth}{August} %% month you are handing in
+\newcommand{\mysubmissionmonth}{February} %% month you are handing in
-\newcommand{\mysubmissionyear}{2020} %% year you are handing in
+\newcommand{\mysubmissionyear}{2023} %% year you are handing in
 \newcommand{\mysubmissiontown}{\myhometown} %% town of handing in (or \myhometown)
 
 %% additional information for generic_documentation title page
@@ -294,8 +294,8 @@
 %\bibliographystyle{abbrv}
 \printbibliography
 
-\appendix                       %% closes main document, appendix follows until end; only available in book-classes
+%\appendix                       %% closes main document, appendix follows until end; only available in book-classes
-\addpart*{Appendix}             %% adding Appendix to tableofcontents
+%\addpart*{Appendix}             %% adding Appendix to tableofcontents
 
 % If the references do not show up, try to run ``biber main'' on the command line and then recompile
 % \printbibliography              %% remove, if using BibTeX instead of biblatex