case-based recommender systems for personalized finance advisory
TRANSCRIPT
Cataldo Musto, Giovanni Semeraro
Case-based Recommender Systems for Personalized Finance Advisory
Graz (Austria) - 16.04.2015
one minute on the Web
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
we can handle 126 bits of information we deal with 393 bits of information
ratio: more than 3x(Source: Adrian C.Ott, The 24-hour customer)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
(from Matrix)
decision-making is actually challenging
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
paradox of choice(Barry Schwartz, TED talk “Why more is less”)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
(financial) overloadC.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
solution: personalizationC.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
to adapt asset portfolios
on the ground of personal user profile and needs
Insight:
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
SolutionRecommender Systems
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Recommender Systems
Relevant items (movies, news, books, etc.) are suggested to the user according to her preferences.
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
definitionRecommender Systems have the goal of guiding the
users in a personalized way to interesting
or useful objects in a large space of possible options.
Burke, 2002 (*)(*) Robin D. Burke: Hybrid Recommender Systems: Survey and Experiments. UMUAI, volume 12, issue 4, 331-370 (2002)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
does it fit our scenario?“we are leaving the age of information, we are entering the age of recommendation”
(C.Anderson, The Long Tail. Wired. October 2004)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Recommender Systems
“[...] The technology is used by shopping websites such as Amazon, which receives about 35 percent of its revenue via product recommendations. It is also used by coupon sites like Groupon; by travel sites to suggest flights, hotels, and rental cars; by social-networking sites such as LinkedIn; by video sites like Netflix to recommend movies and TV shows, and by music, news, and food sites to suggest songs, news stories, and restaurants, respectively. Even financial-services firms recently began using recommender systems to provide alerts for investors about key market events in which they might be interested”
(N.Leavitt, “A technology that comes highly recommended” - http://tinyurl.com/d5y5hyl)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Recommender Systemsfinancial services
http://www.bloomberg.com/news/articles/2015-03-16/smart-beta-etfs-attract-billions-with-critics-blaming-dumb-money
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Recommender Systemssuccess stories
“People who bought…”on Amazon
“Discover”on Spotify
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Recommender SystemsRecommender Systemsunexpected stories
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Recommender SystemsRecommender Systemsunexpected stories
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Recommender SystemsRecommender Systemsunexpected stories
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Recommender SystemsRecommender Systemsunexpected stories
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
recommending financial products is a complex task
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
flocking
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
flocking
Too many users could be moved towards the same suggestions
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
flocking
consequence: price manipulation (as in trader forums)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
poor knowledge
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Features describing both assets classes and private investors are
poorly meaningful
poor knowledge
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
poor history
A combination of asset classes is typically kept for a long time
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Solution
Case-based Recommender SystemsC.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-based RSs• Inspired by case-based reasoning
• Similar problems solved in the past are used as knowledge base
• Reasoning by analogy
• The recommendation process relies on the retrieval and the adaptation of the solutions adopted to solve similar cases
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
....butwhat do we actually mean with ‘case’ ?
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case base
• A case is a the formalization of a previously solved problem
• In our setting
• Description of a user
• Description of a portfolio
• An evaluation of the proposed solution
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-baseexample
user solution evaluation
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-baseexample
user solution evaluation
User Features Risk Profile: LowFinancial Experience: HighFinancial Situation: Very HighInvestment Goals: MediumTemporal Goals: Medium
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-baseexample
user solution evaluation
User Features Risk Profile: LowFinancial Experience: HighFinancial Situation: Very HighInvestment Goals: MediumTemporal Goals: Medium
Euro Bonds 30%
High-Yield Bonds 10%
Fixed-Rate bonds 22%
Euro Stocks 23%
Emerging Market Stocks 7%
Money Market 8%
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-baseexample
user solution evaluation
User Features Risk Profile: LowFinancial Experience: HighFinancial Situation: Very HighInvestment Goals: MediumTemporal Goals: Medium
monthly rate (e.g.)
+0.22%
Euro Bond 30%
High-Yield Bonds 10%
Fixed-Rate bonds 22%
Euro Stocks 23%
Emerging Markets Stocks 7%
Money Market 8%
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-based RSssolving cycle
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-based reasoning for personalized wealth management
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
scenario
“Scrooge McDuck wants to get richer. He decided to invest some of his savings and he asked for help to a
financial advisor”
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
step 1 user modeling
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
scenario
Which features may describe
Scrooge McDuck?
step 1
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
scenario
User Features Risk Profile: Low
Investment Horizon HighInvestment Experience Very High
Investment Goals: MediumFinancial Assets: Medium
step 1
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
User Features Risk Profile: Low
Investment Horizon HighInvestment Experience Very High
Investment Goals: MediumFinancial Assets: Medium
scenario
MiFID-based
step 1
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
scenariostep 1
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
+Generic Demographical Features
User Features Risk Profile: Low
Investment Horizon HighInvestment Experience Very High
Investment Goals: MediumFinancial Assets: Medium
in a classical pipeline, the target user
would have received a “model” portfolio tailored on her profile
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
in a pipeline fostered by a recommender system, the financial advisor can analyze the portfolios proposed to similar users
to tailor the proposal
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
step 2 neighbors identification
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
given a case base, it is necessary to
define a similarity measure to compute how similar two cases are
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
neighbors identificationtrivial similarity: user match
two cases are similar if they share exactly the same features
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
trivial similarity: user match
two cases are similar if they share exactly the same features
neighbors identification
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
neighbors identification
cases are represented as points in a vector space
geometrical alternative: cosine similarity
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
geometrical representationC.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
geometrical alternative: cosine similarityneighbors identification
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-based RSsgeometrical alternative: cosine similarity
each case is seen as a vector
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-based RSsgeometrical alternative: cosine similarity
calculation over the n features
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-based RSsgeometrical alternative: cosine similarity
calculation over the n features
= (risk profile, experience, goals, etc.)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-based RSsgeometrical alternative: cosine similarity
inner product
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-based RSsgeometrical alternative: cosine similarity
it returns the cosine of the angle between A and B
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case-based RSsgeometrical alternative: cosine similarity
case_A
case_B
cosine
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
scenario
case base
step 2
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
scenariostep 2
0.3
0.7
0.9
0.1
similarity score
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
scenariostep 2
0.3
0.7
0.9
0.1
neighborhood(helpful cases)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
step 3 extraction of candidate portfolios
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
scenario
Euro Bonds 30%High Yield Bonds 15%Fixed Rate Bonds 15%
Europe Stocks 20%Emerging Markets Stocks 12%
Money Market 8%
Euro Bonds 30%High Yield Bonds 10%Fixed Rate Bonds 22%
Europe Stocks 23%Emerging Markets Stocks 7%Flessibili Bassa Volatilità 8%
step 2
solutions proposed to the neighbors are labeled as candidate solutions
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
step 4 ranking of candidate portfolios
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
in real-world scenarios, the case base
contains many helpful cases
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
in real-world scenarios, the case base
contains many helpful cases
it is necessary to introduce strategies to filter and rank the cases
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revise
We implemented several ranking strategies
• Temporal ranking
• Clustering
• Diversification
• Financial Confidence Value (FCV)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revisetemporal ranking
solutions are ranked from the newest to the oldest (or viceversa)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Euro Bonds 30%
High Yield Bonds 15%
Fixed Rate Bonds 15%
Europe Stocks 20%
Emerging Markets Stocks 12%
Money Market 8%
Euro Bonds 30%
High Yield Bonds 10%
Fixed Rate Bonds 22%
Europe Stocks 23%
Emerging Markets Stocks 7%
Money Market 8%
Euro Bonds 15%
High Yield Bonds 25%
Fixed Rate Bonds 10%
Europe Stocks 40%
Emerging Markets Stocks 2%
Money Market 8%
Euro Bonds 20%
High Yield Bonds 20%
Fixed Rate Bonds 12%
Europe Stocks 35%
Emerging Markets Stocks 5%
Money Market 8%
revisetemporal ranking
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Euro Bonds 30%
High Yield Bonds 15%
Fixed Rate Bonds 15%
Europe Stocks 20%
Emerging Markets Stocks 12%
Money Market 8%
olderolder
Euro Bonds 30%
High Yield Bonds 10%
Fixed Rate Bonds 22%
Europe Stocks 23%
Emerging Markets Stocks 7%
Money Market 8%
Euro Bonds 15%
High Yield Bonds 25%
Fixed Rate Bonds 10%
Europe Stocks 40%
Emerging Markets Stocks 2%
Money Market 8%
Euro Bonds 20%
High Yield Bonds 20%
Fixed Rate Bonds 12%
Europe Stocks 35%
Emerging Markets Stocks 5%
Money Market 8%
revisetemporal ranking
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
clustering
solutions are clustered and just a small set of centroids is proposed
revise
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
clusteringrevise
Euro Bonds 30%
High Yield Bonds 15%
Fixed Rate Bonds 15%
Europe Stocks 20%
Emerging Markets Stocks 12%
Money Market 8%
Euro Bonds 30%
High Yield Bonds 10%
Fixed Rate Bonds 22%
Europe Stocks 23%
Emerging Markets Stocks 7%
Money Market 8%
Euro Bonds 15%
High Yield Bonds 25%
Fixed Rate Bonds 10%
Europe Stocks 40%
Emerging Markets Stocks 2%
Money Market 8%
Euro Bonds 20%
High Yield Bonds 20%
Fixed Rate Bonds 12%
Europe Stocks 35%
Emerging Markets Stocks 5%
Money Market 8%
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
clusteringcluster 1
reviseEuro Bonds 30%
High Yield Bonds 15%
Fixed Rate Bonds 15%
Europe Stocks 20%
Emerging Markets Stocks 12%
Money Market 8%
Euro Bonds 30%
High Yield Bonds 10%
Fixed Rate Bonds 22%
Europe Stocks 23%
Emerging Markets Stocks 7%
Money Market 8%
Euro Bonds 15%
High Yield Bonds 25%
Fixed Rate Bonds 10%
Europe Stocks 40%
Emerging Markets Stocks 2%
Money Market 8%
Euro Bonds 20%
High Yield Bonds 20%
Fixed Rate Bonds 12%
Europe Stocks 35%
Emerging Markets Stocks 5%
Money Market 8%
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
clusteringcluster 1 cluster 2
reviseEuro Bonds 30%
High Yield Bonds 15%
Fixed Rate Bonds 15%
Europe Stocks 20%
Emerging Markets Stocks 12%
Money Market 8%
Euro Bonds 30%
High Yield Bonds 10%
Fixed Rate Bonds 22%
Europe Stocks 23%
Emerging Markets Stocks 7%
Money Market 8%
Euro Bonds 15%
High Yield Bonds 25%
Fixed Rate Bonds 10%
Europe Stocks 40%
Emerging Markets Stocks 2%
Money Market 8%
Euro Bonds 20%
High Yield Bonds 20%
Fixed Rate Bonds 12%
Europe Stocks 35%
Emerging Markets Stocks 5%
Money Market 8%
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
insight: filtering out too similar solutions
diversification algorithmrevise
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revise
identification of the best subset of similar cases which maximize the relative diversity
diversification algorithm
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revisediversification algorithm
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revise
input similar cases
(candidate solutions)
diversification algorithm
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revise
output subset of
diversified cases
diversification algorithm
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revise
algorithm in each step the portfolio which best diversifies the solutions is
chosen
diversification algorithm
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revise
Solutions with the highest quality are iteratively
chosen
diversification algorithm
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revise
combination between
similarity and diversity
diversification algorithm
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
it returns portfolios that are not so similar to those
previously put in the result set
revisediversification algorithm
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revisediversification algorithm
Euro Bonds 30%
High Yield Bonds 15%
Fixed Rate Bonds 15%
Europe Stocks 20%
Emerging Markets Stocks 12%
Money Market 8%
Euro Bonds 30%
High Yield Bonds 10%
Fixed Rate Bonds 22%
Europe Stocks 23%
Emerging Markets Stocks 7%
Money Market 8%
Euro Bonds 15%
High Yield Bonds 25%
Fixed Rate Bonds 10%
Europe Stocks 40%
Emerging Markets Stocks 2%
Money Market 8%
Euro Bonds 20%
High Yield Bonds 20%
Fixed Rate Bonds 12%
Europe Stocks 35%
Emerging Markets Stocks 5%
Money Market 8%
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revisediversification algorithm
Euro Bonds 30%
High Yield Bonds 15%
Fixed Rate Bonds 15%
Europe Stocks 20%
Emerging Markets Stocks 12%
Money Market 8%
Euro Bonds 30%
High Yield Bonds 10%
Fixed Rate Bonds 22%
Europe Stocks 23%
Emerging Markets Stocks 7%
Money Market 8%
Euro Bonds 15%
High Yield Bonds 25%
Fixed Rate Bonds 10%
Europe Stocks 40%
Emerging Markets Stocks 2%
Money Market 8%
Euro Bonds 20%
High Yield Bonds 20%
Fixed Rate Bonds 12%
Europe Stocks 35%
Emerging Markets Stocks 5%
Money Market 8% XXC.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
reviseFinancial Confidence Value (FCV)
• Simple insight
• We know the historical yield for each of the assets class in the portfolio
• FCV ranks first the solutions composed by a combination of asset classes close to the optimal one (according to previous yield)
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
revise
(Generated yield) (Drift Factor)Total yield is the product of the
yield generated by each asset
class with the its percentage in the
portfolio
Ratio between the yield
generated by the asset classes in the portfolio and its complement
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Financial Confidence Value (FCV)
revise
Euro Bonds - - - 30%
High Yield Bonds 15%
Fixed Rate Bonds 15%
Europe Stocks +++ 20%
Emerging Markets Stocks 12%
Money Market 8%
Euro Bonds - - - 30%
High Yield Bonds 10%
Fixed Rate Bonds 22%
Europe Stocks +++ 23%
Emerging Markets Stocks 7%
Money Market 8%
Euro Bonds - - - 15%
High Yield Bonds 25%
Fixed Rate Bonds 10%
Europe Stocks +++ 40%
Emerging Markets Stocks 2%
Money Market 8%
Euro Bonds - - - 20%
High Yield Bonds 20%
Fixed Rate Bonds 12%
Europe Stocks +++ 35%
Emerging Markets Stocks 5%
Money Market 8%
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Financial Confidence Value (FCV)
revise
Euro Bonds - - - 30%
High Yield Bonds 15%
Fixed Rate Bonds 15%
Europe Stocks +++ 20%
Emerging Markets Stocks 12%
Money Market 8%
Euro Bonds - - - 30%
High Yield Bonds 10%
Fixed Rate Bonds 22%
Europe Stocks +++ 23%
Emerging Markets Stocks 7%
Money Market 8%
Euro Bonds - - - 15%
High Yield Bonds 25%
Fixed Rate Bonds 10%
Europe Stocks +++ 40%
Emerging Markets Stocks 2%
Money Market 8%
Euro Bonds - - - 20%
High Yield Bonds 20%
Fixed Rate Bonds 12%
Europe Stocks +++ 35%
Emerging Markets Stocks 5%
Money Market 8%
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Financial Confidence Value (FCV)
step 5 discussion of the solution
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
financial advisor and private investor
can further discuss the portfolio
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
review
Original Discussed Gap
Euro Bonds 30% 30%High Yield Bonds 12.5% 10% -2.5%Fixed Rate Bonds 18.5% 20% +1.5%
Europe Stocks 21.5% 24% +2.5%Emerging Markets
Stocks 9.5% 8% -1.5%Money Market 8% 8%
interactive personalization
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
step 6 case base update
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
an evaluation score is finally assigned to the proposed solution
yield, e.g.
retain
good solutions are stored in the case base and exploited for future recommendations
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
case base
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
(new) case base
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
our implementation
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
our implementationhttp://193.204.187.192:8080/OBWFinance
demo available
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
OBWFinancelogin screen
advisor-oriented tool
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
OBWFinanceclient selection
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
OBWFinancerecommendation parameters
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
OBWFinanceonly admins can change the parameters
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
OBWFinanceone click to generate recommendations
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
OBWFinancedrop-down menu for selecting the best solution
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
OBWFinanceassets class
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
OBWFinanceyield of the solution
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
OBWFinancechosen portfolio can be further discussed
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
evaluation
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
evaluationwhat is the average yield of
recommended portfolios?
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
evaluationwhat is the average yield of
recommended portfolios?
can recommender systems suggest
better investment portfolios than human advisors?
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
design of the experiment
• 1172 users
• 19 assets classes
• Different neighborhood sizes
• Different features describing the users
• Risk Profile, Investment Goals, Investment Horizon, Investment Experience, Financial Assets, Advice Type, Sex, Age
• Different similarity measures (Cosine vs. UserMatch)
• Leave-one-out experimental design
evaluation
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
experiment 1user match vs. cosine similarity
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Yiel
d
0
0,04
0,08
0,12
0,16
0,2
neighbors
1 5 10
0,20,190,18
0,10,110,09
User Match Cosine Sim
cosine similarity overcomes user match
experiment 2how many features?
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
Yiel
d
0
0,042
0,084
0,126
0,168
0,21
neighbors
1 5 10
0,20,210,2 0,20,190,18
Financial Features Financial + Demographical Features
cosine similarity overcomes user match
experiment 3revise strategies (yield)
best performing configuration provides 0,28% monthly yield
Yiel
d
0
0,056
0,112
0,168
0,224
0,28
neighbors
1 5 10
0,250,240,22
0,270,28
0,220,2
0,150,13 0,14
0,120,09
0,20,210,2
Basic Clustering Diversification FCV FCV + Div
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
experiment 3revise strategies (diversity of the solutions)
ILD=1-average similarity between portfolios
Intra
-Lis
t Div
ersi
ty (I
LD)
0
0,14
0,28
0,42
0,56
0,7
neighbors
0,58
0,35
0,7
0,460,41
Basic Clustering Diversification FCV FCV + Div
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
experiment 4comparison to baselines (leave-one-out evaluation)
recsys better than humans!
Yiel
d
0
0,056
0,112
0,168
0,224
0,28
neighbors
1 5 10
0,270,28
0,220,20,20,2
0,170,170,17
Human Collaborative FCV
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
• FCV calculated on January, 2014
• Recommendations generated on January, 2014
• Evaluation of the yield generated from February 2014 to July 2014
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
experiment 5ex-post evaluation (6 months, with real data)
experiment 5ex-post evaluation (6 months, with real data)
FCV and Diversification is the best one
Yiel
d
0
0,032
0,064
0,096
0,128
0,16
neighbors
1 5 10
0,060,060,060,040,04
0,05
0,110,12
0,16
0,090,1
0,16
0,060,08
0,15
Basic FCV FCV + Div Collaborative Human
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
• Personalized Wealth Management
• Application of case-based reasoning
• Geometrical similarity measure to identify the most similar previously solved cases
• Introduction of diversification and re-ranking techniques
• More than 3% yield for year
• Experiments shows that recommended portfolios overcome the real ones for almost all the users
• Working Demo!
recap
C.Musto, G.Semeraro - Case-based Recommender Systems for Personalized Finance Advisory FinRec 2015 - 1st International Workshop on Personalization and Recommender Systems in Financial Services - Graz (Austria) - 16.04.2015
questions?
Giovanni Semeraro [email protected]
Cataldo Musto [email protected]
in memoriam
Aaron Swartz