console.log('hello');
Welcome to the JavaScript Technology Seminar Projects Show Case
The JavaScript Technology Seminar (JST) is a computer science course given by the rostlab at the Technical University of Munich since 2014. Students of the six weeks JST seminar get hands-on experience developing data-driven applications, working in high-paced agile environment. This page showcases a set of selected projects that students worked on in past semesters.
#JavaScript, #hands-on, #agile, #machine learning, #data science,
#industry experience
During the summer semester of 2016 students of the JavaScript Technology seminar set out to answer a question that was asked by many: Which character is likely to be eliminated during the fifth season of the HBO hit show "Games of Thrones". The students developed applications that scour the web for data about the show and put together a website that reports which characters are most likely to die in the sixth season of the TV series. The project received a world-wide media attention with an estimated reach of 1.2 billion people. In the summer semester of 2019 a follow-up project was run by the JavaScript seminar and received similar attention.
Predict'em All!
Winter2016/17
On July 6th 2016, Niantic released the augmented reality game titled Pokemon Go. The game took the world by storm and is now installed on 5% of smartphones in the US. In the game, players are using their smartphones to locate Pokemons - cute and cuddly virtual creatures that after being captured and nurtured turn into fearless fighters in the name of players' ambition to level up. Locating and capturing Pokemons have quickly become a phenomenon - hoards of people have been sighted in New York as they chase imaginary creatures and try to capture them in Central Park. Pokemons appear on certain places in the real world and wait at those coordinates for a period of time. During the winter semester students came up with an app that predicts a Pokemon’s TLN (Time, Location and Name - that is where Pokemons will appear, at what date and time, and which Pokemon will it be). The app was featured online and managed to predict the appearance of some of the most sought out Pokmeons.
 PokeMapThe PokeMap is screen showing a map with current Pokemon sightings, Pokemon predictions and PokeMobs. It lets users search for a location or a Pokemon, filter by time and/or Pokemon and show a detail sheet for a Pokemon Sighting/Prediction/PokeMob |  PokeDexThe PokeDex shows a list of Pokemons including some of their attributes like name, number and rarity. Additionally, it enables you to search for Pokemon and open a Pokemon PokeDetail page. |  PokeDetailThe PokeDetail page shows all relevant information about a Pokemon, including: Name and number of Pokemon Description, types, weaknesses and strengths Attributes (like weight, flee rate, ...) and evolution Possible attacks And additionally a sentiment analysis based on data extracted from Twitter. |
|---|---|---|
 TeamGuy Yachdav, Tatyana Goldberg, Christian Dallago, Samit Vaidya, Oleksandr Fedotov, Gani Qinami, Paul Gualotuna, Timo Ludwig, Wolfgang Hobmaier, Elma Gazetic, Faris Cakaric, Karen Reyna, Mustafa Kaptan, Georgi Aylov, Benjamin Strobel, Jochen Hartl, Swathi S Sunder, Vivek Sethia, Jonas Heintzenberg, Gilles Tanson, Fabian Buske, Marcel Wagenländer, Annette Köhler, Siamion Karcheuski, Hannes Dorfmann, Alexander Lill, Aurel Roci, Matthias Baur, Timur Khodzhaev, Philippe Buschmann, Josef Brandl |  |
The Music Connection Machine
Summer 2017
Information about classical music is scattered all over the internet in the form of scholarly articles, news stories, blogs, wikis, forums and many other venues. Our goal (in collaboration with Peachnote) is to bring this knowledge into one place and make it easily accessible. We summarize the information about composers, musicians and music works as a set of connections - what were the musicians saying about each other and the music works, and what anybody else has written about them online. Moreover, whenever we find temporal or location-based information, we can present this information in geographic and historic context. The result is a tapestry of information that sums up the knowledge available on the internet about the enchanting world of classical music presented in a fun and interactive way.
 The Music Connection Machine contains data for |  Processing a snapshot of the web* this is equivalent to 1320 hours or 55 days if running on a standard quad core desktop PC |  Filtering rulesFor each website we counted the number of mentioned terms (composer, musician, music piece or instrument). We kept only those websites that mentioned at least 4 and at most 29 different terms. Websites with mentions of 30 or more different terms appeared to be mostly collections and thus were not of interest to us.. |
|---|
TEAM
Guy Yachdav, Tatyana Goldberg, Vlaidmir Viro, Christian Dallago, Kordian Bruck, Phillip Fent, Lukas Navickas, Angelinrashmi Antonyrajan, Tim Henkelmann, Shilpa Gore, Nikita Basargin, Anshul Sharma, Lukas Streit, Felix Schorer, Krishen Kandwal, Hendrik Leppelsack, René Birkeland Birkeland, Anshul Jindal, Daniel Schubert, Sandro Bauer, Lin Ji, Simon Zachau, Martin Mihaylov, Lyubomir Stoykov, Chaoran Chen, Jörn von Henning, Emir Demirdag, Panagiota Revithi, Markus Sosnowski, Yanko Sabev
 Predicted Likelihood of Death<p>The website features a "death predictor" a machine learning-based model that predicts the likelihood of a Game of Thrones character to be eliminated in the latest season of the show. </p> |  Characters Match-UpThe website matches up arch rivals and foretells which of these characters will be the one to drop first |  Machine learning meets pop cultureIn 2016 the Song of Ice and Data app helped bring the concepts behind data science and machine learning to main stream by tying into a pop culture use case |
|---|---|---|
 Facts and StatisticsThe rich world of Ice and Fire is full of (fictional) facts that are just begging to be turned into insights. We present a section that summarized some of the most interesting statistics. Additionally we provide a structured open and freely available dataset. |  A Visual World of Ice and FireThe website features an interactive map that overlays characters, locations and events onto George R.R. Martin fictional universe |  The GoT Twitter SeismographWe automatically scrape twitter for mentions of most of the 2048 GoT characters on Twitter and run a sentiment analysis that gauges audience response. |
 Summer 2016 TeamGuy Yachdav, Christian Dallago, Tatyana Goldberg, Dmitrii Nechaev, Kordian Bruck, Michael Legenc, Sohel Mahmud, Togi Dashnyam, Theodor Cheslerean Boghiu, Boris Idesman\, Georg Gar, Subburam Rajaram, Anna Sesselmann, Nicola De Socio, Thuy Tran, Konstantinos Angelopoulos, Julien Schmidt, Jonas Kaltenbach, Marcus Novotny, Camille Mainz, Santanu Mohanta, Dat Nguyen, Georgi Anastasov, Max Muth, Yasar Kücükkaya, Mina Zaki, Alexander Beischl, Maximilian Bandle, Tobias Piffrader, Florian Gareis |  Summer 2019 TeamGuy Yachdav, Christian Dallago, Ashmin Bhattarai, Fabian Emilius, David Schemm, Gerald Mahlknecht, Daniel Homola,Julian Nalenz, Valentin Dimov, Robert Dillitz, Lukas Franke, Robin Brase, Rainier Klopper, Taylor Lei, Jan Schweizer, Boning Li, Florian Donhauser |
Munich Orbital Verification Experiment (MOVE-II)
Winter 2017/18
MOVE-II is a CubeSat, a tiny satellite with dimensions of 10 x 10 x 13 cm and a mass of 1.2 kg, which was launched into space in December 2018. It is the second satellite of the Technical University of Munich (TUM) and the follow-up project of First-MOVE.
During the winter semester 2017/18 the JST seminar collaborated with the TUM's Scientific Workgroup for Rocketry and Space Flight to build a mission control system with the following objectives in mind:
-
Visualize the satellite
-
Show the satellite’s position
-
Show events like alarms, alerts, commands and data sent and received
-
Provide the operator with the most important information about the vehicle and mission
 Orbit and Altitude VisualizationThe main objective of the Attitude Determination and Control System (ADCS) is to stabilize and to align the satellite towards a defined orientation. We developed a visualization tool that shows the satellite’s position and displays health data / occasional status messages of the ADCS to understand the satellites behavior in orbit and debug the implemented control algorithms. |  3D AnimationThe goal of the 3D model is to visualize the satellite, show the location and purpose of the subsystems, their temperature sensors and the parts that produce the most heat. This model is then used to show data from the subsystem’s live sensor outputs. Additionally this model can be used to show the LEOP animation (solar panels deploying). |  Overview ScreenThe purpose of the Overview Page is to provide the operator with the most important information about the current state of the whole system. |
|---|---|---|
 LogbookThe logbook page shows events like alarms, alerts, commands and data sent and received, etc. This helps the handing over between shifts and enables the operators to react to past events and plan next steps. |  MOVE-II TeamGuy Yachdav, Alexander Lill, Lucie Patzwahl, Maximilian Mumme, Manuel Römer, Xin Yan Yu, Alexander Zillner, Karl Kraus, Dominik Winter, Yixuan Liu, Tobias Klesel, Moritz Schöpf, Rodeina Mohamed, Jonatan Juhas, Florian Mauracher, Marco Grasso, Chen-Hao Chiang, Thomas Zwickl, Riccardo Padovani, Debora Jacoby, Bahareh-Sadat Hosseini |
NLPlot
SUMMER 2018
During the summer term 2018 JST students were building a system (developed in collaboration with Allianz SE) that can translate natural language directives to graphing commands in an agile way. Our prototype receives a set of commands in English and automatically generates visualizations to a given dataset. The system uses then additional commands (again written in English) to manipulate and modify the graphical objects. The system is a first step in designing and building a graphing tool used by professionals who have no expertise in data visualization for their every day business needs. For instance, our system knows how to ingest a dataset from an HR system and show the distribution of employees’ demographics just by taking an input directive such as “plot a histogram of employees age”.
 Two approaches were developed |  Machine learning-based solution |
|---|---|
 Rule-based solution |  Aggregated operationsNLPlot can understand commands that translate into aggregative queries |
 Bar graphsNLPlot understands free text data querying and translates commands into visual objects |  TeamGuy Yachdav, Tatyana Goldberg, Cheng Guo, Roman Priscepov, Hans Kirchner, Shayan Siddiqui, Ahmet Tanakol, Sebastian Stein, Peeranut Chindanonda, Anton Widera, Tobias Priesching, Jyotirmay Senapati, Faisal Hafeez, Shabnam Sadegharmaki, Irakli Tchedia, Carsten Sehlke, Sukanya Raju, Ayishetu Haruna, Jakob Huber |
Software Development Life Cycle Health Predictor
Winter 2018/19
Software development project management success predictor (developed in collaboration with Motius GmbH) - an algorithm that gathers data from various project management systems and predicts whether the project is on track.
TEAM
Guy Yachdav, Alexander Prams, Justin Lübbers, Beatris Burdeva, Martin Rau, Galina Shalygina, Florian Schmid, Maximilian Biber, Simon Kazemi, Sebastian Holler, Julian Ulrich, Julia Dahmen, Jonathan Mengedoht, Frida Gunnarsson, Sebastian Winkler, Bruno Macedo Miguel, David Gogrichiani, Kailiang Dong, Paul Pillau