mdoucem@gmail.com – Page 2

Motivational words for 2018

Posted on January 6, 2018 by mdoucem@gmail.com

I believe good motivational and encouraging words are always needed – to give us that extra push when and if we need it. Thus my first post in 2018 is going to be just about that – some motivational and encouraging words for the year ahead.

The motivational words are not mine but I like them a lot (had posted it here as well). They are from a montage motivational video, called Sacrifice by Les Brown, Eric Thomas and Ray Lewis. I find the words inspirational, and as the video itself does not have subtitles, I have included the words below just in case you are like me and like to have motivational words posted on the wall, work desk or laptop so that you can reflect and visualize them often.

Wishing you all a Happy and Successful 2018!

More women who pitch on Shark Tank get the deal – Shark Tank Analysis Part I

Posted on December 10, 2017December 29, 2017 by mdoucem@gmail.com

Ever since Shark Tank aired in 2009, I have been a fan of the show – from the varied business ideas that get pitched, the background stories of the presenters, to the deals that get made or not made. After several seasons of watching the show, I became curious about certain statistics such as the actual number of females vs males that come to pitch, how many people of color pitch and get deals, what industries gets the most deals, how many of those that got deals shed some tears, etc.

I was thus glad to find a full Shark Tank dataset being kept and maintained by angel investor, Halle Tecco – bit.ly/STankData. With the dataset, I was thus able to explore some of my curiosities. As there are many, I will explore them over two posts. With the first post, I will look at gender distribution, focusing on the eight seasons that are complete. Season 9 is currently ongoing and the analysis below will be updated to reflect that once its completed.

The number crunching was done using R programming, and I have made the R script available on github.

So what did the analysis reveal?

As expected, men constitute the biggest proportion of the presenters:

Looking at all the eight complete seasons, a total of 707 ideas and businesses have been pitched. Of those, 422 were made by Males, 177 by Females, and 108 by a Mixed team.

This trend is apparent from season to season, where we also observe that the show has been having a higher number of seasons since the first three seasons. More specifically, season 2 had the lowest with 36 presenters, while seasons 5, 6, and 7 had the highest with all a total number of 116 presenters each.

Naturally, with the higher proportion of Males in the show, majority of the deals also go to men:

As shown in the figure above, the highest number of females to receive deals were observed in seasons 1 and 2, with 37%. Mixed teams were highest in season 5, at 23% of the total presenters.

Looking at those who got deals, on a good note, more women who pitch get the deal in comparison to males:

Considering just gender specific categories, the data reveals that:

– Out of 177 females that pitched, 99 got the deal (56%)

– Of the the 422 males that pitched, 211 got the deal (50%)

– Similarly to females, 56% of the mixed teams that pitch get a deal (61 out of 108).

The ‘Food and Beverage’ industry has the highest number of presenters:

There are several Industries that are represented on the show and I am always interested to see if there are gender differences in the fields that men and women pitch in. Overall, among the 707 presentations, ‘Food and Beverage’ and ‘Fashion and Beauty’ had the highest representation with 20.8% and 19,8% respectively. The lowest fields represented were ‘Automotive’, ‘Green and CleanTech’, ‘Travel’, and ‘Business and Services’ with 0,99%, 1,41%, 1,41%, and 1,84% respectively.

Notably, Men tend to pitch in ‘Food and Beverage’ and ‘Lifestyle and Home’ industries, while Women are more in ‘Fashion and Beauty’ and ‘Food and Beverage’:

Results reveal that 19,4% (82 of 422) and 16,4% (69 of 422) of males pitched in the ‘Food and Beverage’ and ‘Lifestyle and Home’ industries. While 31,6% (56 of 177) and 22% (39 of 177) females in ‘Fashion and Beauty’ and ‘Food and Beverage’ industries respectively. Similarly, mixed teams were mostly in 24,1% (26 of 108) ‘Food and Beverage’ and 19,4% (21 out of 108) in ‘Fashion and Beauty’.

Notably, no females are recorded to have pitched in the Green and Clean Tech and Automotive fields.

As I work in Tech and constantly hear and observe the low female representation in Tech fields and startups, I was interested to see how this fact is reflected in the Shark Tank data. The data revealed first that the ‘Software and Tech’ industry is fifth in the list of the most presented in industry for men and 8th for women. For some reason, I thought the industry would be higher on the list.

Moreover, of the total females that pitch, 3,39% pitch in the ‘Software and Tech’ industry, while 9,24% of the males pitch in that industry. Roughly half of those who pitched in the Tech field got deals irrespective of gender.

******** That’s it for Part I ********

The above results will be updated to also reflect incoming data for Season 9.

Stay tuned for Part II, where I will explore other factors such as how many people of color have pitched and got deals from season to season.

Guidelines to creating a chatbot

Posted on September 25, 2017September 26, 2017 by mdoucem@gmail.com

With all the hype around chatbots, I was glad to get a chance to actually get some experience in developing one. A few weeks ago, within our Immerstive Automation project, we had Nick Diakopoulos, an Associate Professor in computational and data journalism, lead a workshop on creating chatbots. I was very excited to get to build my first chatbot.

During the workshop, we were introduced to one tool, Chatfuel, for developing Facebook messenger chatbots. It is used by many companies and newsrooms (shown here https://chatfuel.com/bots/) and is easy to use for beginners. Examples of chatbots include those that give you weather updates when you ask, ones that help you pick out and order groceries for the week or news bots that alert you when something interesting happens. For many of these companies, chatbots provide several advantages, as they are able to handle tasks automatically, thus reducing manual hours.

In the workshop, we were teamed up in groups of three. In my group we ended up creating a chatbot that serves you a poem when you ask from it. Nothing fancy as we just implemented a few conversational rules and we didn’t publicly release it. Future plans include experimenting with something more advanced using machine learning, as Chatfuel is rule-based and has limitations in conversational depth.

When designing chatbots, however, here are a few guidelines we were given in the workshop that can guide you on creating a chatbot:

1) Select a specific problem or opportunity that you want to address.

2) Define audience and goal(s) of audience interacting with bot. What will the bot do? Does it solve a problem? Who will use it? Does it serve a particular demographic or niche? In what context will the audience use it?

3) Create a persona for your bot: what are the bot’s goals and behaviors? Will it have a name or personality? What kind of tone will it take?

4) Define the interaction: How will your bot interact? Can you describe a scenario that walks though those interactions step by step? – Good practice here is to try and improvise the conversation: Act out the conversation.

5) Check if there is any data or knowledge-base that your bot needs in order to work?

Lastly, always have a fall back, a user friendly way for your bot to react when it does not ‘understand’ user input.

That’s it! Feel free to share any interesting chatbots that you have created or your favorite tools and approaches, etc.

Six New Publications: in NLG, OSS, and Experimentation

Posted on August 8, 2017 by mdoucem@gmail.com

In academia, they say that you either publish or perish, thus it was good news to know that this year, I was not going to be falling under the latter. In June and July, I received news of six publications that were accepted, putting me in an even better mood before going for summer holidays.

Below is a summary of the papers and their highlights.

Natural language generation

Data-driven news generation for automated journalism: Leppänen, L., Munezero, M., Granroth-Wilding, M. and Toivonen, H., The 10th International Conference on Natural Language Generation (INLG). Santiago de Compostela, Spain, September 2017. To appear.

Highlights: In this paper, we explore the field and challenges associated with building a journalistic natural language generation system. We present a set of requirements that should guide system design, including transparency, accuracy, modifiability and transferability. Guided by the requirements, we present a data-driven architecture for automated journalism that is largely domain and language independent. We illustrate its practical application in the production of news articles upon a user request about the 2017 Finnish municipal elections in three languages, demonstrating the successfulness of the data-driven, modular approach of the design.

Finding and expressing news from structured data: Leppänen, L., Munezero, M., Sirén-Heikel, S., Granroth-Wilding, M. and Toivonen, H., 21st International Academic Mindtrek Conference. Tampere, Finland, September 2017. To appear.

Highlights: This work proposes approaches for automatically finding news or newsworthy events from structured data using statistical analysis and for providing added value to audiences. Utilizing a real natural language news generation system as a case study, we demonstrate the feasibility and benefits of automating those processes. In particular, the paper reveals that through automation of the news generation process, a large amount of news can be expressed in digestible formats to audiences, at varying local levels, and in multiple languages. In addition, automation makes it easier to provide interactivity to the audience allowing them to tailor or personalize the news they want to read.

Open source community analysis

An exploratory analysis of a hybrid OSS company’s forum in search of sales leads: Munezero, M., Kojo, T. and Männistö, T., In Empirical Software Engineering and Measurement (ESEM), 2017 ACM/IEEE International Symposium on. IEEE, Toronto, Canada, November 2017 – To appear

Highlight: The paper presents ongoing work utilizing text analysis techniques to analyze the content of forum posts of a hybrid open source company that offers both free and commercial licenses, in order to help its community manager gain improved understanding of the forum discussions, sentiments, and automatically discover new opportunities such as sales leads.

The many hats and the broken binoculars: State of the practice in developer community management: Mäenpää, H., Munezero, M., Fagerholm, F. and Mikkonen, T., The 13th International Symposium on Open Collaboration (OpenSym), Galway City, Ireland, August 2017 – To appear

Highlight: The paper investigates the varied tasks that community managers perform to ensure the health and vitality of their communities. We describe the challenges managers face while directing the community and seeking support for their work from the analysis tools provided by state-of-the-art software platforms. Our results describe seven roles that community managers may play, highlighting the versatile and people-centric nature of the community manager’s work.

Continuous experimentation

Introducing Continuous Experimentation in Large Software-Intensive Product and Service Organizations: Yaman, S.G., Munezero, M., Münch, J., Fagerholm, F., Syd, O., Aaltola, M., Palmu, C. and Männistö, T., Journal of Systems and Software, 2017 – In Press

Highlights: This article presents a multiple-case study that aims at better understanding the process of introducing continuous experimentation into an organization with an already established development process. Our findings indicate that organizational factors may limit the benefits of experimentation. Moreover, introducing continuous experimentation requires fundamental changes in how companies operate, and a systematic introduction process can increase the chances of a successful start.

Notifying and involving users in experimentation: Ethical perceptions of software practitioners: Yaman, S.G., Fagerholm, F., Munezero, M., Mäenpää, H. and Männistö, T., In Empirical Software Engineering and Measurement (ESEM), 2017 ACM/IEEE International Symposium on. IEEE, Toronto, Canada, November 2017 – To appear

Highlights: This paper examines how ethical issues involved in experimentation are currently understood by practitioners in software development. We conducted a survey within four software companies, inviting employees in different functional roles to indicate their attitudes and perceptions through a number of ethical statements. We observed that employees working in different roles have different viewpoints on ethical issues. While managers are more conscious about company-customer relationships, UX designers appear more familiar with involving users, and developers think that details of experiments can be withheld from users if the results depend on it.

Beginner’s Guide to Doing Text Analysis: Links and Courses

Posted on July 3, 2017 by mdoucem@gmail.com

During the Tech Research Showcase I attended a few weeks ago, someone asked me to guide them to some materials to start learning and doing some natural language processing and text analysis as a beginner. Text analysis, text mining, and natural language processing (NLP) are fields that all deal with mining and analyzing textual data to discover interesting patterns, extract useful insights, or learn more about the structure of language. Sometimes the terms are used interchangeably, even though the fields have different goals and work at different levels of language analysis. In addition, they all include techniques from linguistics, mathematics, statistics, and computer science.

In this post, I will not go into introducing the fields as they has been covered in several resources, and frankly, they are too huge to cover well in one post. They include a vast amount of techniques and methods for performing language detection, translation, document summarization, sentiment analysis, part-of-speech tagging, topic modeling, question answering, etc. A good background book which is practically oriented is the Practical Text Mining and Statistical Analysis for Non-structured Text Data Applications.

As large fields that have gained increased interest in both Academia and Industry, there are several sources and ways one can start. However, for someone who wants to just get their feet wet with text analysis, I would recommend the following links and courses to get practically started. From these, you also get to understand what is involved and helps you identify what questions or what insights you might want to get from the analysis or mining of the text, e.g., what is the sentiment in the text? What is the genre? Is it spam? among many others.

Recommendations:

-Natural Language Processing with Python: Analyzing Text with the Natural Language Toolkit – a practical introduction to NLP. You will learn by example, write real programs, and grasp the value of being able to test an idea through implementation.

–Python for text analysis – A practical course in Python, geared towards those who want to get some hands-on experience working with language data.

Apart from python based, there is also R based practical courses that make use of the tm: Text Mining Package, for example, Introduction to Text Mining with R for Information Professionals.

For online course which provide both theory and practical materials

–Introduction to Natural Language Processing – Provides an introduction to the field of NLP. The programming assignments are in Python.

–Text Mining and Analytics – The course will cover the major techniques for mining and analyzing text data to discover interesting patterns, extract useful knowledge, and support decision making, with an emphasis on statistical approaches that can be generally applied to arbitrary text data in any natural language with no or minimum human effort.

Besides understanding the field itself, recently it has become important to also understand how machine learning models are powering NLP applications. Recently, deep learning approaches have obtained very high performance across many different NLP tasks.

–Deep Learning for Natural Language Processing – Online course with lectures and reading materials – Lecture videos are available here.

–Oxford Deep NLP 2017 course – An advanced course on natural language processing containing lecture slides.

I will continue updating this page as I come across better or additional materials. If you have other good suggestions for beginners, please feel free to share.

There should be more research pitching events

Posted on June 21, 2017 by mdoucem@gmail.com

Last week, I took part in my first Tech Research Showcase. As a researcher who is interested in entrepreneurship I wondered afterwards why I had never attended a pitching event where the sole focus was my research. The tech research showcase event was organised by Icebreaker.vc at the startup hub Maria 01.

Icebreaker invited tech researchers from around Finland to come and give a five minute pitch on their research to an audience of business people, developers and designers. The event was motivated by the world-class research conducted in Finland and by the several companies that have started as a result of it, for example, Relex, Beddit, and Callstats.io.

I was attracted to the event, first because one could pitch their research even if you were at the state where no commercialization value had been identified – the aim of the event was to put researchers in a room with experienced people that have the ability to collaborate, point you in the right direction or give some business advice. And since commercialization is something that has come up in my current work project Immersive Automation, I wanted to gauge the interest and potential of our research. Secondly, I was challenged by presenting my research in five minutes. Normally, I have done research presentations in conferences, seminars, lectures, etc., events where I have anywhere between 20 to 1 hour, significantly more time. But 5 minutes, in front of people who might not know the field was a challenge that I wanted to undertake.

For the pitch itself, we as the presenters were advised by the organizers to follow a format where we presented the following:

The problem that is being solved – put in a way that it could be understood by a non-expert.
How the problem is solved today. What is the state-of-the-art?
What it is you have discovered in your research. What is the new idea?
What the results are using your method compared to the best solution today.
What you are looking for. For example, particular skills, co-founders, or guidance. This should be as concrete as possible.

In total we were 12 researchers pitching, with majority of the presenters being from Aalto University and University of Helsinki. Machine learning, as might be expected was the dominating theme among the presentations and the presentations ranged from ideas to already launched startups.

Having participated in a few business pitching events, one advice I could give to fellow researchers, myself included, would be to participate in more training on selling ones research work and ideas. A few presentations went into too much detail about the technicalities of the research such that it was difficult for some of the audience members to fully grasp the potential of the idea. A good training course that is focused on how to sell research work and ideas is the one provided by the brothers Andy and Steve Langdon. The two-day course is practically oriented, one learns a lot of selling techniques, and its organised in a fun and relaxed way.

Actually, I think many universities should internally organize events like these as a way for researchers to learn and practice selling their work, a skill I believe is needed in today’s competitive world. I found having to think and prepare a five minute pitch interesting and it does make you look at your research from various other angles besides research publications.

Continuous experimentation for personal development

Posted on May 22, 2017 by mdoucem@gmail.com

A few weeks ago, my colleagues and I published a book titled “Continuous Experimentation Cookbook: An introduction to systematic experimentation for software-intensive businesses”. The book is a result of several years of work within the Empirical Software Engineering research group and with software development teams in medium and large companies that were part of the Need for Speed project. The book provides an introduction to continuous experimentation, which is a systematic approach to continuously testing product or service value or whether a business strategy is moving in the right direction.

Continuous experimentation as a development and decision-making approach has always been one that I found not only to apply to software or businesses development. It could actually also be applied to personal development. Just like in business and product or service development, experimentation, through small, fast, and cheap probes, can allow us to gain deeper insight of what we really want, what makes us happy or what works for us in terms of achieving our goals.

As the cookbook lays out, to conduct successful experiments, a few key elements should be understood (captured in Figure 1). Although these elements are specific to software development in the book, I will use a personal development example to illustrate how this approach might be applicable (see Table 1).

Figure 1: The continuous experimentation cycle that links decision-making to systematic experimentation (Adapted from [1]).

Table 1: Key elements in continuous experimentation applied to a personal development example (Example narrative is in Orange color).

Notably, not all the experimentation elements that are presented in the cookbook or in Table 1 will be necessary for every personal development experiment. In addition, I think that for personal development experiments, there is more flexibility on how scientific they should be. For instance, in experimentation, one must introduce the change and then hold everything else constant to observe cause and effect, but in life it is hard to hold things constant and this might interfere with the results. But I believe a clear hypothesis and good execution plan can mitigate those factors, and can help you make informed decisions where you need them or just want to be certain.

To conclude, continuous experimentation, just as with software or business development, can also work for our personal development, by helping us test our assumptions about what we want or think is true about ourselves, and then help us make informed decisions and changes. Of course this approach is one of many other approaches to decision-making. As highlighted in the cookbook, one of the pitfalls to avoid is making decisions solely from the collected data. As humans, we still have our intuition and experiences to guide the decision making process (see discussion of this in a previous post). But of course decision-making is guided by having access to factual information.

Would definitely love to hear your experiences on how or where you would apply this approach to your personal development.

Source:

[1] Munezero, M., Yaman G. S., Fagerholm F., Kettunen, P., Mäenpää H., et al.: Continuous Experimentation Cookbook: An Introduction to Systematic Experimentation for Software-intensive Businesses, DIMECC Publications Series No.15, Helsinki, Finland, 2017.

Credit:

Post edited by Fabian Fagerholm – https://www.linkedin.com/in/fabianfagerholm

Balancing privacy and ethical rights when detecting harmful behaviors

Posted on May 2, 2017December 29, 2017 by mdoucem@gmail.com

Over the past few years, companies like Google, Twitter, or Facebook have had to walk a fine line between policing negative behaviors such as hate speech, terrorism, fake news, etc., and supporting basic rights of free speech online. With the increase in aggressive, terrorism, cyberbullying, hate speech, and racist communication, as well as expressions of depression or suicidal thoughts on social media, there are however increased pressures and incentives on these organizations and researchers (like myself) to develop fully or semi-automated methods to detect these negative behaviors. However, as pointed out in my dissertation, balancing the prevention of these harmful behaviors with privacy and ethical rights is a challenge, but this conversation is something that should be had.

In my PhD research, for instance, I aimed at building antisocial behavior detection models that could be incorporated into real-world systems and act as early warning systems for security enforcement organizations or institutions like schools. Although the research respected privacy rights, with these types of detection there is always the concern over the access and analysis of personal information, with concerns that it might be just be a step away from ‘big brother’.

Facebook, with its almost two billion users and use of sophisticated algorithms, privacy and ethical rights are always a question mark. This year, for example, it introduced a semi-automated depression detection algorithm that analyzes and spot patterns of posts that are potential suicidal, and then sends them to the Facebook team for appropriate response [1, 2]. Earlier, this identification process used to be only manual.

It is true that a lot of data is public or that users on certain platforms give away the rights to their data. This makes it easier to develop these detection algorithms, however, it might not be the wish of a Facebook or twitter user to have their data analyzed for purposes that they have not given their consent to. This was for instance observed when it was reported in June 2014 [3] that Facebook was using status updates to manipulate users’ moods and observe how that manipulation translated to their status updates. Even though with Facebook one does agree to their data policy when using their application, people however are often not aware of what is being done with their data and it is hard to draw the line of whether it is alright or if it is infringing on privacy rights. Though, an argument could be made that we must just accept that anything that is posted on social media platforms is not private.

Ethical and privacy issues do arouse real concerns that have an impact on the broad areas of developing and using detection algorithms to detect harmful behaviors. If data is public or accessible, should researchers or organizations still ask for consent to use the data for research or development purposes? If yes, to whom should the consent be acquired from? With so much available data, it is the task of researchers and practitioners to make sure that the data acquired is only used for good. But how do we measure what is good and what might be considered invasive or infringement of free speech? It might be argued that if data and technology are used to prevent crimes or improve quality of life, then it should be allowed. However, under those reasons, organizations can have the excuse or reason to automatically monitor and prevent the posting or sharing of messages that are deemed harmful from their perspective. With such policing, one might wonder if uprisings like those in Egypt in 2011 which made use of social media to organize, schedule, and spread the uprisings would have been possible [4, 5].

It is hard to see what a good solution to this would be, but perhaps to take advice from Ray Kurzweil [6], maybe the answer to these ethical and privacy concerns is to have a set of standards that are established through a whole social discussion between technologists and society, within and across different societies.

References:

[1] Matt Burgess, (6 March 2017), How tech giants are using AI to prevent self-harm and suicide, Wired, http://www.wired.co.uk/article/facebook-safety-self-harm-suicide-ai-instagram, (visited on 2017-04-27).

[1] Natt Garun, (1 March 2017), Facebook leverages artificial intelligence for suicide prevention, The Verge, http://www.theverge.com/2017/3/1/14779120/facebook-suicide-prevention-tool-artificial-intelligence-live-messenger, (visited on 2017-04-27).

[3] Kashmir Hill (28 June 2014), Facebook manipulated 689, 003 users’ emotions for science, Forbes, https://www.forbes.com/sites/kashmirhill/2014/06/28/facebook-manipulated-689003-users-emotions-for-science/#4736d97197c5, (visited on 2017-04-27).

[4] Erick Schonfeld, (16 February 2011), The Egyptian behind #Jan25: “Twitter is a very important tool for protesters”, TechCruch, https://techcrunch.com/2011/02/16/jan25-twitter-egypt/ (visited on 2017-04-29)

[5] Sam Gustin, (11 February 2011), Social media sparked, accelerated Egypt’s revolutionary fire, Wired, https://www.wired.com/2011/02/egypts-revolutionary-fire/ (visited on 2017-04-29)

[6] Bill Joy and Ray Kurzweil, (12 July 2001), Future shock: High technology and the human aspect, Hoover Institution, http://www.hoover.org/research/future-shock-high-technology-and-human-prospect (visited on 2017-04-27).

Detecting antisocial behavior in text

Posted on April 20, 2017December 29, 2017 by mdoucem@gmail.com

The words we use and our writing styles can reveal information about our preferences, thoughts, emotions and intentions. Using this information, I developed machine learning models that can detect antisocial behaviors, such as hate speech and indications of violence, from texts, as part of my recently defended PhD dissertation, titled “Leveraging emotion and word based features for antisocial behavior detection in user-generated content.”

Historically, most attempts to address antisocial behavior have been done from educational, social and psychological points of view. My PhD research, however, demonstrated the potential of using natural language processing techniques to develop state-of-the-art solutions to detect antisocial behavior in written communication.

The research created solutions that can be integrated in web forums or social media websites to automatically or semi-automatically detect potential incidences of antisocial behavior with high accuracy, allowing for fast and reliable warnings and interventions to be made before the possible acts of violence are committed.

One of the great challenges in detecting antisocial behavior is first defining what precisely counts as antisocial behavior and then determining how to detect such phenomena. Thus, using an exploratory and interdisciplinary approach, I applied natural language processing techniques to identify, extract, and utilize the linguistic features, including emotional features, pertaining to antisocial behavior.

The research investigated emotions and their role or presence in antisocial behavior. Literature in the fields of psychology and cognitive science shows that emotions have a direct or indirect role in instigating antisocial behavior. Thus, for the analysis of emotions in written language, the research created a novel resource for analyzing emotions. This resource further contributes to sub-fields of natural language processing, such as emotion and sentiment analysis.

Because a problem in researching antisocial behavior in written language was that there was no adequate collection of texts, the research, in addition, created a novel corpus of antisocial behavior texts. The corpus allowed and will continue to allow for gaining deeper insight and understanding of how antisocial behavior is expressed in written language.

The study showed that natural language processing techniques can help detect antisocial behavior, which is a step towards its prevention in society. With continued research on the relationships between natural language and societal concerns and with a multidisciplinary effort in building automated means to assess the probability of harmful behavior, much progress can be made.

Doctoral dissertation is available for download at: http://epublications.uef.fi/pub/urn_isbn_978-952-61-2464-3/index_en.html

In the press:

Hilary Lamb (13th April 2017), Computers taught to recognise hate speech and violent language, Engineering and Technology, https://eandt.theiet.org/content/articles/2017/04/computers-taught-to-recognise-hate-speech-and-violent-language/

University of Eastern Finland (12 April 2017), New machine learning models can detect hate speech, violence from texts, ScienceDaily, www.sciencedaily.com/releases/2017/04/170412091222.htm

Terhi Nevalainen, (11th April 2017), Tietokone voi tunnistaa terroristin, Karjalainen, http://www.karjalainen.fi/uutiset/uutis-alueet/kotimaa/item/138639-tietokone-voi-tunnistaa-terroristin

Trump speech generation using Markoc chains

Posted on February 26, 2017December 29, 2017 by mdoucem@gmail.com

Moving from Trump speech analysis (see previous post) to speech generation, in this post I investigate the possibilities of automatically generating text that could plausibly have been by Trump. For the generation, I experimented with using Markov chains. Markov chains, named after Andrey Markov, are stateless mathematical models that describe sequences of possible events. They are used in several real world applications such as autocomplete suggestions, speech recognition, text identifiers, path recognition, and in many other artificial intelligence tools.

A Markov chain tells the probability of transitioning from one state to any other state. In text generation, Markov chains learn the probability of the next word or character given one, two, or more words or characters. In this post, I will not delve into the details and mathematical formulation of Markov chains. For a visual description of how Markov chains work, see Victor Power’s site.

As this was my first practical experiment working with Markov chains, I thus began by using python code written by Nevo [1], which I then slightly modified for the purpose of generating Trump like speeches.

I started by creating a collection of Trump’s speeches. The collection included words from his inauguration speech and his first and final presidential debates. In total, the text collection consisted of 16,667 words. I first trained a Markov chain model on uni-grams (one word), then bigrams (two words) and trigrams (three words) and continuously increased the n-gram order. Then the model was given a random word or words to begin from, and it started generating the rest. I limited the length of generated text to 150 words.

Below are the results. What do you think, do the following texts sound like a speech made by Trump?

A: Text generated from model learned with unigrams

“I am a brilliant community, you tried to happen. And what do not show that. It’s an airplane on to that. I mean, I told them $1,500, and peaceful transfer of state when you right around election time, including a small group came out about young and that’s true throughout the true way you find this? Is it was so bad. First of our country. And I am going to Aleppo? OK, so many different reality exists: mothers and I should have taken the war in every year. You haven’t even worse, and I can’t do damage to be obsolete, because of the likes of them to have led our country, whether a routine audit. And I’ll go forward to myself, “I can’t bring into the clips where so much tougher and look — now, and other country, you would be a disaster.”

B: Text generated from model learned with bigrams

“We share one heart, one home, and one glorious destiny. The oath of allegiance to the manufacturing industry. You go to Mexico and other countries, while allowing for the presidency of the people she wants to make America great again. Thank you. Today’s ceremony, however, has very special meaning because today, we are not keeping up with the nations of the greatest assets anywhere in the world, beyond the United States got along well, that would help you, I would like to… It’s not an accurate one at all. It’s not an accurate one at all. It’s not nice. And I think I’ve ever seen in any government whatsoever. Wrong. CLINTON: … it is… Lester, how much? Typical politician. All talk, no action. Sounds good, doesn’t work. Never going to create new deals.”

C: Text generated from model learned with trigrams

“We, the citizens of America, are now joined in a great national effort to rebuild our country and our economy generally is — look, we owe $20 trillion. We cannot do it any longer, Lester. HOLT: Back to the question, though. How do you bring back — specifically bring back jobs, American manufacturers? How do you miss $6 billion? You ran the State Department, $6 billion was missing. How do you bring back — specifically bring back jobs, American manufacturers? How do you miss $6 billion? You ran the State Department, $6 billion was missing. How do you make them bring the jobs back? Well, the first thing you do is don’t let the jobs leave. The companies are leaving. I could name, I mean, there are thousands of them. They’re leaving, and they’re leaving in bigger numbers than ever.”

As can be seen, the text generated by the model trained with unigrams (A) is nonsensical. With the bigrams trained model (B), the text starts getting a bit better, but still does not make much sense. With the trigrams trained model (C), the text begins to sound like some thing Trump would say.

From my experiments, trigrams were identified as the highest possible order before the generated text resulted in direct quotations of the text used in the training and thus the models were no longer generating text per se. I also tried generating longer texts of more than 150 words, but it resulted in nonsensical texts. This is because the Markov chain models do no have a memory beyond the set n-grams. Thus, new sentences might be generated with topics that are not related to prior sentences.

Based on this experiment, I would say that Markov chains are not the best approach in generating speeches of considerable length, and future experiments involve identifying better approaches. For short texts, like Tweets, I have observed good results being achieved by recurrent neural networks. As an example, a Trump Twitterbot @DeepDrumpf produces quiet Donald Trump sounding tweets, for instance the following tweet “America has never been more harmed by the vote. I made a lot of money on that. I am doing big jobs in places, now everything is Benghazi.” Thus, a most likely next experiment will be to experiment with recurrent neural networks for generating longer texts.

Source:

[1] Omer Nevo (2016). Poetry in Python: Using Markov Chains to generate texts. https://www.youtube.com/watch?v=-51qWZdA8zM

Trump’s usage of adjectives and adverbs

Posted on February 25, 2017December 29, 2017 by mdoucem@gmail.com

By now, many of us have heard President Donald Trump speeches, or at least snippets of them. One thing I have noticed, among many other things, is that he tends to use a lot adjectives and adverbs or at least I always get the notion that they are many. Most likely, it could be just that it’s the same short adjectives and adverbs that are repeated over and over and thus sounds like he uses them a lot, for example in the following phrases; “…build a very huge wall”; “It’s going to be really great“; “so sad, tremendously sad, greatest sadness ever.”

Since the overuse of adjectives and adverbs can be seen as embellishing and can clutter sentences pointlessly, especially in formal speeches, I was curious about how Trump actually uses adjectives and adverbs in comparison to for instance the former president Obama, whose speeches have been said to be more eloquent.

To investigate, I compared transcripts of the inauguration speeches by Trump (2017) and by Obama (2009, 2013), and their first news press conference as president elect (Trump in 2017 and Obama in 2008). The press conference speeches included the president elects’ answers to posed questions by reporters. Table below summarizes the number of words in each speech. From the table, we can observe that there was a greater range in the number of words used by Trump in his first news press conference and inauguration speech, while Obama’s three speeches are relatively about the same in the number of words.

Table: Total number of words in presidential inauguration speeches.

For the analysis of the adjectives and adverbs, I made use of TreeTagger. TreeTagger is a tool for annotating text with part-of-speech tags. Part-of-speech (POS) tagging is the process of marking up words in a text with their part of speech, e.g., noun, verb, adjective, adverb, etc. After performing the part-of-speech tagging, I retrieved for each speech, only the word-POS pairs, where the POS tag was an adjective or adverb. From the retrieved list, I performed a comparative analysis of the usage of the adjectives and adverbs between Trump and Obama in their inauguration speeches (A) and in their first news press conference as president elect (B).

A. Inauguration speeches

The figures below show the distribution of adjectives and adverbs in each inauguration speech, as a percentage of the total words in the speech. From the figures, we can see that there is a higher use of adjectives and adverbs in Trump’s inauguration speech than in both Obama’s 09 and 13 speeches. Interestingly from the figures, we can see that Obama’s use of adjectives and adverbs has relatively been the same across both of his speeches.

From the results alone, it is difficult to judge whether they indicate an under, normal or over-usage of adjectives and adverbs. Thus, to get some indication, I included in the figures a LIWC mean score for both the adjectives and adverbs. The LIWC mean score was obtained from the popular Linguistic Inquirer Word Count (LIWC) text analysis tool. The tool includes a dictionary of words built from analyzing over 100,00 files of text , representing over 250 million words. In building the dictionary, it was identified that on average, adjectives constituted 4.49% of the total words and adverbs 5.27%. Thus, from the figures we can observe that while both Trump and Obama used significantly a higher number of adjectives than the LIWC mean, with adverbs, Obama is around the average, while Trump is visibly above.

Unfortunately, since majority of the text files used in developing the LIWC tool were not political type of speeches, a bigger comparison with other speeches to identify what is ‘average’ in the political context and where Trump’s speeches fall, would need to be conducted.

Exactly which adjectives and adverbs does Trump uses, was the focus of my next analysis. Figure below reveals the top 20 adjectives and adverbs that were most frequently us by Trump in his inauguration speech. Using those same 20 words, I identified how frequently they appeared in Obama’s speeches. The results are also shown in the figure.

From the figure, we can see Trump’s speech had high usage of the words “Americans,” “again,” “back,” and “great.” This is reflective of his inauguration speech, which was America-centric and was focused on making America great again and bringing things back to Americans.

Surprisingly one of the most frequent term in the three inauguration speeches is the adverb “not”, with Obama using it more often in his speeches than Trump. For example, it was used in Obama’s 13 speech in the following phrases: “our journey is ‘not’ complete until our wives…”; “‘not’ complete until our gay brothers…”; “‘not’ complete until no citizen…”

The adverbs “new” and “now” were also emphasized in all the inauguration speeches, perhaps indicating the presidents’ desires to bring in new things now or bring in change as presidents.

B. News press conference speeches

Moving a bit back in time from the inauguration speeches to the then president elects’ first news press conference, I analyzed how they used adjectives and adverbs. The analysis revealed that in his press conference, Trump used 5,95% adjectives while Obama used 6,97% and Trump used 7,79% adverbs while Obama used 5,61%.

In addition, I also looked at the top ten most frequent adjectives and adverbs. These are shown in the figures below.

From the figures, we can see the different usage of adjectives and adverbs. In particular, the adverb “very” is used significantly more by Trump than Obama. It was used by Trump in phrases such as “I am going to work very hard”; “I’m very proud..”; ” I look very much forward”; “… going to have a very, very elegant day.” From the press conference results, we can also see that there is a difference in the length of adjectives and adverbs used. Specifically, the average length of the adjectives and adverbs used by Trump is 4 characters while for Obama it is 5,8 characters.

In summary, this analysis has mostly served to reveal the actual usage of adjectives and adverbs in Trump’s speeches. It is interesting to see for instance the change in the top adjectives and adverbs used by Trump from the press conference to the inauguration speech. Notably, the adverb “very” was used significantly less in the inauguration speech. In addition, we can see that from all the speeches analyzed, there was a tendency for Trump to use more adverbs than adjectives when compared to Obama. However, due to the small sample of speeches analyzed, it is not possible to make any conclusive deductions. Further studies will need to be conducted.

Sources:

Trump’s 2017 inauguration transcript – https://www.washingtonpost.com/news/the-fix/wp/2017/01/20/donald-trumps-full-inauguration-speech-transcript-annotated/?utm_term=.7c244dd73119

Trump’s 2017 news press conference transcript – https://www.nytimes.com/2017/01/11/us/politics/trump-press-conference-transcript.html?_r=0

Obama’s 2013 inauguration transcript – https://www.theguardian.com/world/2013/jan/21/barack-obama-2013-inaugural-address

Obama’s 2009 inauguration transcript – http://abcnews.go.com/Politics/Inauguration/president-obama-inauguration-speech-transcript/story?id=6689022

Obama’s 2008 news press conference transcript – http://www.washingtonpost.com/wp-dyn/content/article/2009/03/24/AR2009032403036.html

Any change in diversity with the 2017 Oscar nominations?

Posted on January 29, 2017December 29, 2017 by mdoucem@gmail.com

When the 2015 and 2016 Academy Award nominations were released, many in Hollywood and on social media were deeply offended by the lack of racial diversity among the nominees, especially in the prominent categories of best actor / actress, and best supporting actor / actress, where only white actors and actresses were nominated. So did any changes take place in 2016 that are reflected in the 2017 nomination list?, especially the representation of people of color in the nomination list?

After the 2015 nomination announcement, the #OscarSoWhite and #OscarNorms hashtags began trending on Twitter as that was the first since 1998 that no person of color, Hispanic or Asian was nominated for the Academy Awards in the acting categories. Moreover, when the 2016 nominations were released and again the acting categories were only white people, the #OscarSoWhite hashtag resurfaced, with an outcry on social media and a boycott of the Oscars spearheaded by Jada Pinkett-Smith and Spike Lee. The outcry after the 2016 nominations was also a result from many feeling that 2015 had produced material that was worthy of nominations and that the Academy had passed over well-reviewed performances in the movies Creed, Straight Outta Compton and overlooked prominent actors of color like Idris Elba (Beasts of No Nation), Michael B. Jordan (Creed), Will Smith (in Concussion), and the many young actors in “Straight Outta Compton”. And even when there was nominations for these movies, only white people were nominated. For example, for the movie Creed, only Sylvester Stallone was nominated, while the film’s black writer-director, Ryan Coogler as well as the lead actor, Michael B. Jordan, were not nominated.

Because of the upset from the 2016 nominations, Academy president Cheryl Boone Issacs ushered in new membership rules and added 683 new members as a way to diversify a predominantly white, male and elderly group. The academy now numbers 6,687 people. These are the first Oscars voted since this change [source].

So did 2016 bring in new opportunities for diversity at the Oscars? The 2017 Oscar nominations were released last week on the 24th January and in this post I analyze the changes, if any, that have occurred in this year’s nominations in comparison to those of 2015 and 2016. For this post, I only analyze the nominees in the acting categories as well as the directing categories, this is because many on social media also felt that there were directors of color that were passed over by the Academy, for example Ava DuAvernay for her work in Selma.

Table below summarizes the nominees for the 2017 Oscars in the four acting categories.

In comparing the above nominees to previous two years, the Figure below shows the ethnic composition among the nominees in the four acting categories. The figure captures the distribution of white people (blue), people of color (orange) and other ethnicities – ‘other’ (grey).

From the figure above we see that there has been a positive change in the number of nominations for people of color in 2017 when compared to 2015 and 2016. Especially in the best supporting actress category, where majority of the actress nominated are black.

I further looked at the changes, if any, among the nominated directors. These include directors nominated in the Best Director category and the directors of the documentaries included in the Best Documentary Feature category. Figure below gives the distribution of white people (blue), people of color (orange) and other ethnicities (grey) in the years 2015 to 2017.

From the figure above, we see that the 2017 nomination list had a higher representation of persons of color than 2015 and 2016, especially for the Best Documentary Feature category. Notably, four out of the five directors in the nominated feature documentary are persons of color: Ava DuVernay, Raoul Peck, Roger Ross Williams, and Ezra Edelman. In particular, among the nominated directors for both film and documentaries, there is still a dominance of males, with only two females in 2015 (Laura Poitras and Rory Kennedy), one in 2016 (Liz Garbus) and one in 2017 (Ava DuVernay), all being in the Documentary Feature category. Ava DuVernay is actually the only director who is female and a person of color among the analyzed three years.

So, has there been more diversity among the 2017 nominees? Based on the analysis above, this year’s nominations definitely have a higher representation of persons of color in the acting and director categories than the previous two years. With a total increase from zero nominees in 2015 and 2016 to 11 nominees in 2017 (see Figure below). In particular, the biggest change is reflected among the best supporting actress and best feature documentary categories. However, There is still a small number of other ethinicities represented, even in the 2017 nominations.

There is no reason however to see these changes as due to the social media outcry or the boycott. Many of these movies have been years in the making and all the actors and actress are on those lists due to their own merits. More good quality movies directed, produced and acted by a diverse set of people will lead to more diversity in the Oscar nominations.

(P.s: If you notice any mistakes in the data above please let me know.)

Using content imagery experiments to increase Netflix viewership

Posted on October 31, 2016October 31, 2016 by mdoucem@gmail.com

If you are like me, you might have experienced long tiresome moments when browsing and searching for something interesting to watch on Netflix. Majority of this browsing is usually constant scrolling up and down through the images of the content until I find something that catches my eye. When I do, I usually read the story of the title, the actors and rating to see if its worth a try. But often, there is nothing that catches my interest and I quit or settle on an old Friend’s episode.

As a data collecting savvy company, I wondered what Netflix did with this user behavior and whether it performed any experiments to see how the imagery affects the content watched.

In following up with this wondering, I stumbled upon two articles from Netflix, detailing the experiments they do with content imagery as a way to identify ways to improve viewership. Netflix knows that with all the large amount of content they provide, they have a short time to capture the attention and interest of users. Particularly, since the human brain can process an image in as little as 13 milliseconds [1]. In addition, they understand that imagery of their content is the most efficient and compelling way to do that [2]. In one of their consumer research studies, done in 2014, they found that the imagery was the biggest influencer to a user’s decision to watch content, and that it also constituted over 82% of users focus while browsing the content. They also found that users spent an average of 1.8 seconds considering each content that is available on Netflix, which is a very, very short time.

Knowing this, Netflix thus conducts several experiments to try and find a ‘right’ imagery that captures the attention and interest of the users in that short amount of time. The experiments are usually run using A/B testing (explained here). During the experiments (detailed more here), Netflix collects various measurements such as click through rate, aggregate play duration, fraction of plays with short duration, fraction of content viewed, etc. [3].

In this post, however, I would like to share some of their key learnings from conducting imagery experiments as a way to improve their service offering [2].

1. Images showing faces with complex emotions outperform stoic or benign expressions. They identified that seeing a range of emotions actually compels people to watch a content more. This could be because complex emotions have a better ability to convey a large amount information regarding the tone or feel of the content. This was observed for instance in their testing of a ‘right’ image of the second season of Unbreakable Kimmy Schmidt.

2. Regional differences still exist, and are important for some content and imagery. A good example of when they observed the importance of regional presentation of a content and how it can impact its discovery among users around the world, was with the Sense8 TV Show. Sense8 has an international cast and storylines that give it a diverse appeal and makes it resonate with varying types of people. Thus, when developing the imagery for Sense8, this diversity was reflected in the final images, showing how much they varied between different countries and cultures.

3. Personally, I have thought that villains more than heroes make or destroy a movie, especially in the case of action movies. Thus it was not surprising to me that one of Netflix’s finding was that using visible, recognizable characters (and especially polarizing ones) results in more engagement. In particular, Netflix found that their users respond to villainous characters surprisingly well in both kids and action genres. For instance, in Dragons: Race to the Edge, the two images of villainous characters seen below significantly outperformed all others.

4. Three is apparently the upper limit of cast size one can show in the imagery. This is particularly for small sized artwork where a large cast size is not as effective in helping users decide to play a content. For huge billboards this might be the opposite though. During experimentation, Netflix observed this preference when they saw a drop tendency when an imagery contained more than three people. This finding directly informed their imagery for Orange is the New Black.

Season 1

orange-s1

Season 2

orange-s2

Season 3

orange-s3

Based on the experiments, it is clear to Netflix that using better images to represent content significantly increased overall streaming hours and engagement from their members [2]. Many of the above four findings are intuitively known by many of us Netflix users, but it is always interesting to read what companies are doing to improve their offering and product experience.

Sources:

[1] Trafton, Anne (16-01-2014), In the Blink of an Eye. MIT News. Retrieved on 23^rd July 2016 http://news.mit.edu/2014/in-the-blink-of-an-eye-0116

[2] Nelson, Nick (03-05-2016). The Power of a Picture. Netflix Media Center. Retrieved on 26^th June 2016, https://media.netflix.com/en/company-blog/the-power-of-a-picture

[3] Krishnan, Gopal (03-05-2016), Selecting the Best Artwork for Videos through A/B Testing. The Netflix Tech Blog, 26^th June 2016, http://techblog.netflix.com/2016/05/selecting-best-artwork-for-videos.html

Keep data as a tool and the brain as the decision maker

Posted on October 10, 2016October 29, 2016 by mdoucem@gmail.com

Having been working with in the area of data and text analysis, and experiment-driven software development for a couple of years now, combined with my occasional enjoyment of well-made TV shows, I naturally found myself listening to Sebastian Wernicke’s TED talk to the finish. The talk is about using data to make (smart) decisions. I found the talk quiet relevant in today’s world of big data, and increased availability and accessibility to software usage data. Big data has rapidly moved into many real-life decision making processes in the workplace, law enforcement, medicine, etc., where serious decisions are being driven or aided by data.

What I liked from Wernicke’s TED talk titled ‘How to use data to make a hit TV Show’ was that it was a reminder that even though access to huge data has opened up many opportunities in various fields, data is still just a tool and decisions should not be solely driven by it. Wernicke emphasizes that the thing between our ears, i.e., brain (taking into account that that thing has the expertise to make sense of what the data is saying), should be the driver of decision-making.

In the talk, Wernicke gives two examples from two very competitive and data-savvy companies (Amazon and Netflix), who both collect and analyze millions of data points from their customers/users, to illustrate his point. In one example, big data was used successfully (in the case of Netflix with the House of Cards TV show) and in another example, not so successfully (case of Amazon in their creation of the Alpha House TV show).

As Wernicke explains, when Amazon wanted to create a TV show, they started by taking various ideas from people. From those ideas, they selected eight TV show candidates. Then they made a first episode of each one of these eight shows and put them online for free for the public to watch. Then Roy Price (Head of Amazon Studios) and the team at Amazon recorded everything, i.e., from when somebody pressed play, pressed pause, what parts they skipped, what parts they repeated, etc. They collected millions of data points because they want to use those data points to then decide which shows they should make. They did all the data crunching from those millions of collected data points and an answer emerged. And the answer in this case was that “Amazon should do a sitcom about four Republican US Senators”, and they made that show. They used the data to drive their decision making and ended up with a show that was not so successful – ‘Alpha House’ (Alpha House has an IMDB score of: 7.6).

This not so successful case of Alpha House was compared by Wernicke to the more successful story of Netflix in their creation of House of Cards (House of Cards has an IMDB score of: 9.0). Netflix approach was to start by looking at all the data they already had bout their viewers such as the viewer ratings, viewing history, and so on. Then they used that data to explore and discover little bits and pieces about the audience: what kinds of shows the viewers liked, the producers they liked, the actors, etc. With all these pieces collected, they took a risk and decided to license a drama series bout a single senator which was ‘House of Cards’.

Wernicke uses these two examples to explain the difference in how data can be used to make decisions. In the talk, he continues to explain that whenever we, as humans, are solving complex problems, we are essentially doing two things: The first is to break that problem into bits and pieces so that we can deeply analyze each of those bits and pieces. The second part involves then putting all these bits and pieces back together to come to our solution – and sometimes, this is an iterative process. Data and data analysis are only good for the first part, that is, no matter how powerful that data and data analysis is, it can only help us in taking a problem apart and understanding its pieces. It’s not suited to putting those pieces back together again and then to come to a conclusion. Wernicke points out that there is another tool that can put the pieces back together, and its available to all of all of us, this tool of course is the brain, and one of the things it is good at, is taking those bits and pieces back together again, even when there is incomplete information, and coming to a good conclusion. But to come to a good conclusion, that brain has to have some expertise.

Of course data helps us see what we might miss, find new avenues or business ideas. Thus in our data analysis work, we must get the balance right. Yes, “more data is better and can deliver brilliant insights, but in the end it has to be integrated by expert human brains for complex issues like producing a brilliant TV show”, Wernicke states. Humans as decision makers still need to be part of the data analysis equation, but what this also means for data scientists is also to have the expertise to make ‘often’ right conclusions or risks, Wernicke adds. The sentiment of the brain as the decision maker was also echoed by Beverly Wright, executive director at the business analytics center at Georgia Tech, in a Keynote Panel Discussion at Global Big Data Conference [Source].

From the two examples that Wernicke gives, making the right conclusions often as a data scientist or knowing when to take risks, comes with practice. For instance, from the time when those two TV show examples were created, both companies have learned a lot since then. For example, this year, 2016, Amazon had two original series nominated for the golden globe awards.